This is a good question with many possible answers and points-of-view. It seems that it is often the case that we learn about human behaviour indirectly by considering how humans interact with other humans, the environment and machines. In short, mathematical views of human behaviour often focus on stimulus-response modelling.
Mathematical models of human behaviour have been recently been studied in the context of epidemics in
P. Poletti, Human behaviour in epidemic modelling, Ph.D. thesis, University of Trento, 2010:
http://eprints-phd.biblio.unitn.it/422/1/tesi.pdf
Human behaviour is modelled by Poletti in terms of two mutually influencing phenomena: epidemic transitions and behavioural changes in the population of susceptible individuals (see Section 2.2, starting on page 16).
Modelling human-computer (device) interaction is the focus of
P. Eslambolchilar, Making sense of interaction using a model-based approach, Ph.D. thesis, National University of Ireland, Maynooth, 2006:
http://www.dcs.gla.ac.uk/~rod/publications/Esl07-thesisFinal.pdf
See, for example, the probabilistic framework of a model-based behaviour system in Fig. 6.13, starting on page 182.
A bit less mathematical but still very interesting model of human behaviour in terms of human-made music is given in
A. Tidemann, A groovy virtual drummer: Learning by imitation using a self-organizing connectionist architecture, Ph.D. thesis, Norwegian University of Science and Technology, 209:
http://www.idi.ntnu.no/research/doctor_theses/tidemann.pdf
For an overview of the Tidemann's approach to model and imitate human musical expressiveness, see Fig. 1.1, p. 5.
This thesis introduces the SHEILA architecture in terms of human drum-playing patterns with an accompanying melody (see Section 3, Architectuure, page 106 in the pdf file but unnumbered in the thesis).
You can also use game theory applications for modelling human behavior ..
Camerer, C.F.: Behavioral Game Theory: Experiments in Strategic Interaction. (eBook and Hardcover)http://press.princeton.edu/titles/7517.html
Though it may not directly be relevant , it will definitely offer insight and direction. He has also published the tables and results of various studies undertaken on behavior.
I think one of the approaches is the use of Bayesian inference system or Bayesian model....fuzzy logic and fuzzy theory can be used to describe human behavior in mathematical terms
What characters are you using as you dependent variables? Define the variables that constitute the behavior you want to model.
Dear All,
This is a complicated issue. I am a layman in this field but I would like to know the reliability of similar models and the behavioural patterns where models like those can be used. Regarding James’s example, epidemic models can be important and useful considering the danger of Ebola transmission in Africa.
Dear All,
Down-voting Abedallah’s answer is but very scarce information and cannot clear the issue.
There could be so large number of possible independent and dependent variables (human behavior) and countless possible mathematical equations from simplex algorithms to simple regression to multiple regression. Human behavior is that complex.
The most difficult parts of this is generating the data and calculating the "coefficients" to build the mathematical models for prediction.
This is one of those great open-ended questions to which there could be a thousand - valid and more or less operationalizable answers.
James Peters answer is a great set of places for you to start.
If one is speaking of adoption/adaptation of innovation, then Everett Rogers classic The Diffusion of Innovations now in its Fifth Edition is the best place to start in my view, for a relatively straightforward and yet potentially highly informative modelling approach. The diffusion methodology is derived from epedemiology and is the starting point for -all- new product marketing and sales forecasting models and methods, whether they credit Rogers or not.
On other hand, if we are talking social media modeling, then the theory of the wisdom of crowds has more to do with - animal swarm behavior modeling than modeling humans making conscious choices... or you could be better off thinking of this as an artificially intelligent agent problem, in which you are focusing again on modeling a limited set of definable behaviors....see Sebastian Thrun's AI textbook; he is now best known for his MOOC/Udacity co-founder fame. (Although so far Udacity shows he understands machines better than human learning behaviors....in my humble opinion ; )
And speaking of learning and education, there are of course a wide variety of methods for modeling these human behaviors as well as variety of approaches to technology-augmented learning. At: https://www.researchgate.net/profile/Lee_McKnight/contributions?ev=prf_act you can find a variety of articles related to 'collaborative learning' and 'interactive learning' which could be helpful if that is part of your intended model.
Of course, there are whole fields of study such as human factors, psychology, sociology, anthropology, political science, and economics to name a few which could offer you a wide range of models and approaches to consider as well; which if any is what you seek we cannot tell from your brief question.
Basic microeconomics is all about modeling human behavior in markets and hierarchies, which can lead one to models of transaction costs, production costs, and so on...so for you any Micro 101 textbook may be the place you should start, as we can only guess from your wide-open ended question : )
Returning to business and technology - focused modeling approaches which can readily incorporate various aspects of the above disciplines into a complex system of systems model, the 'systems dynamics' modeling approach mastered at MIT's Sloan School of Management and a handful of other educational institutions can be incredibly illuminating; on other hand speaking as thesis advisor to several students who did doctoral studies employing systems dynamics models, it is highly labor intensive to do quality work. But if not done to focus of a serious doctoral thesis study, is otherwise prone to sloppy garbage in - garbage out glaring conceptual errors. Although still offering a false impression of rigor and precision because of the model's complexity. (Of course, MY students all did great and careful jobs! ; )
Last and likely least, our own work on 'Coordinating User and Device Behaviors' may offer you some possible thoughts and approaches; again depending upon what one is trying to accomplish. If what you are thinking of is relevant to cognitive radios and dynamic wireless networks and smart devices...and Internet of Things and clouds...and humans interacting across networks and devices....then check it out in revised form as as a chapter in the Fitzek and Katz Springer book 'Cognitive Wireless Networks' (2007) or you can find an earlier version also in my Researchgate.net publication list that was published by CMU in 2003.
Good luck! (And maybe tell us more precisely about what you are trying to do/what you did? : ) Since it is unclear what precisely you are attempting to do, no doubt some of the links below are off-target; but perhaps one or two may save you some time and hunting. For those in US, best wishes for a restful Thanksgiving; for our Canadian cousins, a belated happy Thanksgiving; and to everyone else, best wishes for a productive week!!!
https://www.researchgate.net/profile/Lee_McKnight/contributions?ev=prf_act
Article Interactive Learning through Wireless Grids
Conference Paper Emergent learning through playful interactions and serious g...
Data A Failure to Communicate: Security Vulnerabilities in the Gr...
Chapter Coordinating User and Device Behavior in Wireless Grids
Article Wireless Grids: Assessing a New Technology from a User Perspective
Article Content pricing in the Internet
Article Show Me the Money: Contracts and Agents in the Service Level...
Dear Morteza,
Pentland and Liu (1999) described a set of dynamic models (e.g., Kalman filters) sequenced together by a Markov chain to model and predict human behavior. The paper is attached. I have not gone through the previous answers yet and don't know if it was mentioned before, so I hope it is relevant to your question.
Pentland and Liu (1999) Modeling and Prediction of Human Behavior. Neural Computation 11, 229–242.
Behavior results from a symbolic black box (e.g. millions of chemical reactions) of which the content/process is probably inaccessible to mathematical models that wish to predict with high precision the future?
Can mathematicians model and predict their own behavior, also given they have access to information concerning environment-individual interactions?
It think it would be wise to model some aspect of human behaviour. This can result in good agreement with actual behaviour. Modeling a complete human behaviour is more than a super comuter job !
The use of mathematical models in the experimental analysis of behavior has increased over the years, and they offer several advantages. Mathematical models require theorists to be precise and unambiguous, often allowing comparisons of competing theories that sound similar when stated in words. Sometimes different mathematical models may make equally accurate predictions for a large body of data. In such cases, it is important to find and investigate situations for which the competing models make different predictions because, unless two models are actually mathematically equivalent, they are based on different assumptions about the psychological processes that underlie an observed behavior. Mathematical models developed in basic behavioral research have been used to predict and control behavior in applied settings, and they have guided research in other areas of psychology
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1472627/
As we humans are complex, our behaviors in most cases are intent and hidden goal driven. But fuzzy mathematics (a multivalued logical structure ) can at least help to start an incomplete study of few of our behaviors that are less involved in our thoughtful decision process.
As James put it eloquently these few behaviors of humans can be studied via stimuli-response mechanism in fuzzy logic setting, but the problem in this mechanism is that the responses might not be the actual or real reactions conceived by the person from the stimuli but an a priori intent driven response ( in most cases a thoughtful decision ) for a posteriori purpose.
Such study that bases on stimuli-response processes on human behaviors might not be therefore a real mathematical study but purely an entertainment. For instance a school principal wants to know how many kids feel happy today and those who are not happy will receive some unknown gifts so that they will be happy for the whole day.
As happiness is a feeling that can not be measured or detected by some happy-o-meter but an information gained from direct responses of individuals, the answer for this question may be true or may not be true. Some kids will simply say, they are not happy today (although they might actually are as happy as ever), because they want to know what the unknown gift will be and want to get one. Therefore in this case the study fails in principle to get a real response of a stimuli.
Although I brought school kids as an example but true to almost all humans who habitually make decisions based on thinking and presumed a posteriori results or intents on any stimuli they encountered.
I believe human behavior can be modeled. It is complex, but definitely amenable to a few rules.
There are already work going on in Analytic in modelling human behavior. it is not perfect yet, but we are getting there. While I would definitely agree to the argument that it may not be 100% accurate, but we can be reasonably sure most of the occasions.
There are many Psychological tools with quantification assessment. A very complex behaviour like predicting user choices; based on what the user did in the past, given that to predict what they will be doing in the future, requires more knowledge about the target person, it is very much possible to predict behaviour through scientific process translated and generalised using maths.
http://web4.cs.ucl.ac.uk/staff/T.Jambor/blog/2010/10/29/to-what-extent-mathematical-models-can-mimic-human-behaviour/
It is supposed that game theory is an approach, but it is probably an overestimated approach, since it cannot give you the time dynamics of the behavior: It says that, "look this is the final equilibrium stage", but when will this be reached? Nobody knows...
Of course human behavior can be modeled: in dealing with the world in our own actions we are always predicting, with some level of confidence, the reactions of others. In most cases we do not need person-by-person predictions, but estimated distributions of group responses. This is particularly clear for economic settings (both in macroeconomics and for management, as marketing), for epidemiology, for traffic flow, and various other settings where we note significant use of such models. The difficulty lies in constructing, at least for some restricted class of behaviors, models which are simple enough to work with and yet correspond reasonably with reality --- just as for models of physical problems in mechanics or chemistry or ... . A cautionary point is the use of optimization models in economics (i.e., that each individual always acts to maximize utility) which do not accurately reflect the way actual humans think and behave --- which is, e.g., why a psychiatrist Kahnemann got the Nobel Prize in economics. My conclusion is that the modeling of human behavior is difficult so there is much very bad modeling in the literature, but that does not preclude the possibility of good behavioral modeling and I do think the situation is improving.
Until you come out with a theory, testable and falsifiable of human behavior, your efforts to model it are worthless.
Dear All,
I think typical often stereotype behavioural patterns can be modelled. That is models may help mainly at population level. I mean people whose cultural and psychical backgrounds and environmental conditions are similar. As to individuals, in very typical, society or situation determined cases like very quick or very slow processes (escaping from a danger or choosing food items or way of life) models can help. I think marketing specialists may use models like those.
It deepens upon the kind of problem you want to study.
maybe evacuation behaviour, driving behaviour, ..., marketing behaviour erc.
Could you please tell more about your problem?
I think the question brings more practical life problems. To develop a model over human behavior more studies over cultural background, rituals, attitudes , family history etc.are essential. To do this, a test using powerful questionnaire may be applied upon several population sample. Scoring them. Use factor analysis, derive respective communality and specificity function. Then develop a model. Thank u all.
Why Dear Yaron Rosenstein, maybe he lacks think of parameters that are related to the non-physical side of the human being? !!
The Science continues, and it will succeed one day !
Look at this modeling:
http://www.youtube.com/watch?v=H2vN2QXZGnc
This seems to be a question interesting to many of us, but until we hear again from Morteza Noruzi with a somewhat more specific description of the type of problem it is impossible to have any idea as to what kinds of suggestions might actually be useful to him.
Dear Noureddine-Ouerfelli, how do you model non-physical parameters ?
Does it even make sense?
I do agree with Prof. Ouerfelli, and that video is very telling.
In fact here is the paperback for the same title since the video may or may not be used as a reference in a paper (Something to be discussed in a separate question I am sure).
http://www.amazon.com/Mathematics-Sex-Clio-Cresswell/dp/1741141591
May be the attached interview can shed more light into the questions and respond to some the concerns raised including those of Yaron .
Kindest regards,
FBMB.
Really ?
A video ?
An interview ?
Are you people even familiar with scientific facts gathering ?
What a person says is not scientific evidence.
Analysis of short wildlife films : a case study for mathematical modeling?
A film lasting about four minutes on YouTube might illustrate examples of wildlife behaviour, like plant-animal interactions during pollination. The film can practically be subdivided into a large number of pictures on a screen of 1 m². Each individual picture of the film might also be subdivided into smaller or larger compartments. If one picture would be subdivided into a puzzle of 10.000 cm² and each 1 cm² is visualized and analysed separately, which part of the screen is required to get an idea about the information provided by the whole screen and the whole film? Doing this might result in a continuous flow of an uncountable number of potential hypotheses. Different people watching the whole film, a fraction of the film, or a fraction of the screen presenting one picture of the film will produce different hypotheses based on what they perceive and what they memorized in the past. Different potential hypotheses will definitely follow different probability distributions depending on who is watching the film and scales of analysis or perception involved. The probability that two observers will produce exactly the same hypothesis using the same wording and terminology will probably be very low. How many hypotheses might be formulated focusing on only 1 cm² from a screen of 1 m² and what will be the scientific approach used? Hypotheses and methods used to test them will definitely change with changes in scales of analysis and perception, e.g. going from 1 pixel to 1 cm² to 2 cm² up to 1m². Pixel analysis might focus on physics of colours whereas analyses of whole pictures might focus on physics of colours or colour contrasts but also on other aspects, like how and why two filmed species interact. Moreover, how many mathematical equations will be required to describe film details concerned with pollination of flowers? If people only would have access to Mathematics describing the film they did not see before, how would they translate equations into film pictures? How would different mathematicians cope with 20 pages of mathematical equations to be transformed into visual pictures of which contents become accessible to citizens? All this illustrates that interactions between methods and hypothesis-testing involving people varying in background knowledge might potentially result in an uncountable number of hypotheses and related conclusions, evidently depending on scales of analysis or perception involved. It also implies that the same problem might be tackled in many different ways depending on who will start and conduct research, probably related to past biology-based experience of the researchers involved?
How do you decide what aspect of behavior should be mathematically modeled given that there are a large number of options available?
A well-known example that relates behavior to mathematical modelling is Phenology or proxies of Phenology (e.g. timing patterns related to reproduction). One of the reasons why phenology has been studied frequently might be that it is not so dynamic at the time of sampling/measurement and that it expresses large variation at different spatial-temporal scales.
Dear Marcel,
I may be wrong but I thought the subject was modelling human behaviour?
Dear Andras,
The topic of the wildlife film could be researchers conducting a project or researchers as observers of non-human behavior (see above)
Science in practice and biodiversity: can science behavior (e.g. behavior of scientists willing to initiate a project) be mathematically modeled?
What is the role of founder effects in science practice, e.g. depending on the background knowledge of the researchers that start a project you will end-up with different predictions/different methods/different results, etc.... Who decides who will initiate a research project (the financial organization willing to find sexy results, the director of the laboratory, the researcher proposing a project, the research environment of the researcher proposing a project, etc...)?
Can the mathematical analysis/modeling of science behavior reliably predict the outcome of a research project? An advantage is that you can ask researchers why they made a behavioral decision, which is not possible or more complicated with non-human organisms.
Dear Yaron,
We are playfully discussing. Most of us left the academic dress at home.
Dear Marcel,
I am not against wildlife films representing animals and plants. Your answer was beautifully formed and regarding its complexity one could apply it as s/he wishes. Or I am not open sufficiently for the diversity of opportunities and approaches.
Indeed Andras. And how will you mathematically model these behavioral science decisions with predictive power?
Marcel, human behavior is too complex, many parameters involved, I just don't see how you can model it deterministically. I.e. A finite set of odes
Why should human behavior differ from non-human vertebrate behavior in relation to mathematical modeling?
The use of mathematical models in the experimental analysis of behavior has increased over the years, and they offer several advantages, as they require theorists to be precise and unambiguous, often allowing comparisons of competing theories that sound similar when stated in words. Hence, a good mathematical model can provide a common framework for understanding what might otherwise appear to be diverse and unrelated behavioral phenomena.
Mathematical models of behavior can be used to predict or control behavior in applied settings. Such models are used [1] in neuroscience and psychopharmacology to help researchers identify the functions of different brain structures and to assess the behavioral effects of different drugs.
[1] Platt M.L, Glimcher P.W. Neural correlates of decision variables in the parietal cortex. Nature. 1999;400:233–238.
Let's take the behavior of the human modeler. S/he can decide to behave following or against the predictions or the mathematical model, or not. Perhaps behavior that result from unconscious-based mechanisms can be mathematically modeled better than conscious-based mechanisms?
Continue with my own line of reasoning! Psychologists vary in their quantitative skills and in their tolerance for mathematical equations, it is important for those who develop mathematical models of behavior to find ways (such as verbal analogies, pictorial representations, or concrete examples) to communicate the key premises of their models to nonspecialists. However, for some psychologists and behavior analysts who began their careers when mathematical modeling was not so commonplace in this field, or for those who do not use mathematical models in their own work, this trend may be disconcerting. For some, the sight of an equation or two in an article may be reason enough for them to skip over the article and move on to the next. They may feel that an article with mathematical equations is beyond their comprehension, or worse, irrelevant to their interests. After all, isn't the experimental analysis of behavior supposed to be about behavior, not about mathematical equations and symbols?
I used the excerpt from the following article (attached) in this post.
James E Mazur. Mathematical Models and the Experimental Analysis of Behavior, J Exp Anal Behav. Mar 2006; 85(2): 275–291. doi: 10.1901/jeab.2006.65-05
Continue with my own line of reasoning: Do mathematical modelers of human behavior take their own experiences into account to produce the models, e.g. that they accept that their own behavior can be used as a reference of a whole target population?
No. they are just approximation but still practical in medicine and other places. Some models of human behavior in the social sciences assume that humans can be reasonably approximated or described as "rational" entities. As I said before, mathematics is not a goal or end in itself, rather a tool in such studies, so long as it can contribute to the effort. The purpose is not simply to create “mathematical models” for their own sake. What they are good for actually? Mathematical models can be used to further our understanding of the factors that motivate individuals in , for example, making medical decisions, taking into account the interplay among biological systems, individual decision-making and social and cultural influences
You can not define it uniquely, you can not measure it.
That is why you can not model it.
The line between behavioural economics and psychology is blurred. Behavioural economics is based on the traditional “neoclassical” model of human behaviour used by economists. This essentially mathematical model says human decisions can usefully be modelled as though our choices were the outcome of solving differential equations.
But these models ignore the underlying biology-based mechanisms that occur within the individuals.
Example:
When two individuals A and B are exposed to the sample stimulus A, the two individuals will not respond in the same way to stimulus A because they differ in underlying biology-based mechanisms that occur within the individuals (inaccessible mechanism/black box A interacting with stimulus A to cause behavior A versus inaccessible mechanism/ black box B interacting with stimulus A to cause behavior B).
@ Marcel
Biology is not only thing in human beings, more of "feelings and emotions" play in behaviour modeling.
The emotions/feelings result from what is happening inside the individual. According to neuro-endocrinologists there exist inside the individual body axes/pathways that connect external/internal stimuli, with brain components and/or organs to influence behavioral expression. Are these neuro-endocrine axes with all their individual-based characteristics accessible to modelers?
Psychology, economics, sociology etc. are not sciences.
They have no clear definitions, observables and no central testable theories.
Biology as well, i'm not sure it's a science.
Biology is descriptive not explanatory.
Dear All. Are not we going too fast? I think psychology, economics, sociology, etc ARE sciences and they all belong to social sciences
For example, sociology is the science or study of the origin, development, organization, and functioning of human society; the science of the fundamental laws of social relationships, institutions, etc. he meaning of the word comes from the suffix "-logy" which means "study of," derived from Greek, and the stem "socio-" which is from the Latin word socius, meaning member, friend, or ally, thus referring to people in general. It is a social science involving the study of the social lives of people, groups, and societies, sometimes defined as the study of social interactions. It is a relatively new academic discipline which evolved in the early 19th century.
Same can be argued about Biology, etc.
Dear Mahmoud, do you have central testable theories in the social sciences?
No.
Can you conduct doubly-blind controlled experiments in the social sciences ?
For ethical reasons, big-no.
A short answer to both questions is YES, but I am an engineer so I better leave the discussion and prove to the expertise in social sciences. Also, I fully support my previous answer
i finish this with John Lubbock quote: 'What we do see depends mainly on what we look for.
Yaron, every science has its own theory (or story). If we dont find them to fit to our theories (i.e., central testable theories or what so ever) It doe not mean they are not based on sound ideas. A central theme is whether one scientific discipline can be reduced to the terms of another. All I know that there is no clear theory on many questions or central problems in science, including if at all science can reveal the truth about unobservable things or scientific reasoning can be justified at all. Besides, some consider problems that apply to particular sciences such as biology or physics. Some may use contemporary results in science to reach to conclusions .
Human behavior in an evolutionary framework
Origin and maintenance of human-created artificial objects
The exact reasons for the proximate origin of biodiversity cannot be revealed simply because we cannot ask non-human organisms about the timing and types of selection or creation processes. Perhaps the role of natural or sexual selection processes that create and maintain phenotypic diversity can be more easily identified in human-created domesticated living beings or even human-created artificial objects, such as the evolution and diversity of cars, toys, houses, or paintings.
+ Can human created artificial objects (e.g. the shape or colour of cars) be studied to investigate underlying mechanisms creating phylogenetic or phylocultural trees and its link with selection or creation processes? For instance, people creating artificial objects could be asked why they created these objects and buyers (receivers) selecting these objects could be asked why they selected or rejected these objects? Perhaps this would be an adequate and practical approach to study and understand underlying mechanisms of adaptation and selection, resembling approaches of those that study adaptation and selection in artificial test tubes in standardised laboratory conditions.
+ To what extent do human-created objects and living beings share the same principles in design and phenotypic diversity?
+ What selection forces are responsible for the spread of human-created objects, for instance the amount of cars sold?
+ Is the creation of a new car, like the creation of a new individual phenotype, a random or deterministic process?
+ Why do human created artificial objects and living beings differ in structural complexity? Because of human-intellectual constraints or functional constraints (limited function equals limited design)? For instance, mechanics do not have sufficient biological background knowledge or biologists do not have sufficient background knowledge in mechanics retarding development of artificial objects with biological characteristics. For instance, car design is simple when only used to drive with (tools influencing speed and direction on prepared roads). For instance, living beings are more complex than human-created artificial objects because of significant differences between the time a project was initiated and the phenotype was delivered.
+ Accepting phenotype is much simpler for human created artificial objects (e.g. a car) than for living beings, why is the opposite observed for houses and nests. For instance, human-created houses and their contents are considered to be much more complex than animal constructions used for living and nesting, such as a bird nest in simple box?
Human-created artificial odours and the application of evolutionary-based rules
What is the impact of human-created artificial odours in social mediation? Are odours more important than words in negotiations?
+ Do man-created artificial odours hide information, for instance from a biological point of view? For instance, are human-created artificial odours potential biological traps, or not? If odours are involved in individual recognition or recognition of individual phenotypes, ‘who is who’ in a world involving human-created artificial odours?
+ Did human-created artificial odours influence the social structure of whole societies, and if so, why? For instance, do human-produced artificial perfumes create new combinations of human biodiversity absent in the absence of artificial odours, and what happens with social interactions if human-created odours are removed?
+ If strong odours block olfaction, what is perceived, or not?
+ Are those that create artificial odours representative for the huge diversity in human olfaction systems at the population level?
+ Preferred or non-preferred odours are they culture or physiology dependent?
+ Why do some human-created odours have more success than others?
+ What is the function of multi-cocktail odours environments at meetings or airports? For instance, are cocktails of human-created artificial odours examples of multi-therapy mediating human-non-human interactions in places where many people come together, for instance in relation to risks of diseases transmissions?
+ Do human-created odours avoid infection or do they contribute to development of new micro-organism strains as in antibiotics?
+ How does wildlife (or domesticated animals with sensitive olfaction systems, like dogs) respond to human-created odours, and what are biological consequences, for instance from a social or life-history point of view?
+ Why do odours used by wildlife, such as the use of medicinal plants in blue tits, have lower concentrations than human-created odours applied for human social interactions? For instance, do strong human-created artificial odours overstimulate the immune system (e.g. allergy), or not?
+ Experiments involving the impact of human-created artificial odours on wildlife, do they have to be conducted by medical faculties (e.g. INSERM) or non-medical faculties (e.g. INRA, CNRS), but what about the scientific background knowledge of such experiments required to improve both experimental design and data interpretation abilities, such as former experience with wildlife?
+ How can medical specialists treat diseases (e.g. viruses, bacteria) they are not working on, for instance in the framework of vaccination campaigns? Do they just follow recommendations from pharmaceutical companies, and if so, what is the difference of scientific competence between medical doctors (e.g. 13 years of education) and medical assistants (e.g. 3 years of education) in these circumstances?
What is the scientific definition of perfumes: research on the reliability of publicity?
According to law, publicity should be reliable and therefore use terminologies that are reliable. Is this always the case from a scientific point of view?
Example
The liquid in a bottle can be named a perfume, but what are the criteria to decide whether the liquid is a perfume, or not? If the same liquid is perceived differently according to perceivers or perceived differently by the same individual according to the treatment of the liquid with the environment (air, skin), is the definition mainly based on the emotional responses of perceivers, or based on well-defined chemical compositions of the liquids? For instance, can any odour be defined as a perfume if the reaction of the perceiver is excitement, surprise, or what else? If the definition of perfume is mainly linked to chemical composition, and chemical composition can influence biological reactions, why are chemical details not (always) provided on bottles or boxes making publicity, as is the case for food?
Let me be more specific about my last answer. Yaron believes "Psychology, economics, sociology, biology, etc. are not sciences." If so, my question is What are they!
Economics! The reason I selected economics from above list is clear, Economics most closely resembles the natural sciences, of all the social sciences . Economic theories have been axiomatized, and articles and books of economics are full of theorems. Let's see what economics is . It is about inquiries concerning [1].
(a) rational choice,
(b) the appraisal of economic outcomes, institutions and processes, and
(c) the ontology of economic phenomena and the possibilities of acquiring knowledge of them.
These inquiries lead into three subject matters:
Economic theories of rationality, welfare, and social choice defend substantive philosophical theses often informed by relevant philosophical literature and of evident interest to those interested in action theory, philosophical psychology, and social and political philosophy. Also, economics is of particular interest to those interested in epistemology and philosophy of science both because of its detailed peculiarities and because it possesses many of the overt features of the natural sciences, while its object consists of social phenomena.
Economics is a test case for those concerned with the extent of the similarities between the natural and social sciences [1]. Those who have wondered whether social sciences must differ fundamentally from the natural sciences argue that "owing to the importance of human choices (or perhaps free will), are social phenomena too “irregular” to be captured within a framework of laws and theories?" Given human free will, perhaps human behavior is intrinsically unpredictable and not subject to any laws. But there are, in fact, many regularities in human action, and given the enormous causal complexity characterizing some natural systems, the natural sciences must cope with many irregularities, too.
[1] Philosophy of Economics, Link is provided
http://plato.stanford.edu/entries/economics/
To me science is not constrained to modelling, which seems to be the usual frame of mind of most har scientists I met. They think ev erything can be modelled. It may be the case, but models only represent simplified realities. Therefore I am inclined to say that some aspects of human behaviour can be modelled, but not the entire behaviour.
The reason being that human undergo constant qualitative change which means that mathematics as a language used in modelling cannot really tells us much!
Mahmoud, every science relates one way or the other
To a physical theory.
Social sciences do not relate to physical theories.
Hence psychology is not a science.
Yaron, I am hopeless to continue any further in this issue. I think this issue is very important and it needs better attention and fuller discussion. This can be cast into a new question by you or other members..
We are moving away from the original question by making Morteza more confused to find an answer to his question of how he can mathematically model the behavior of the humans. So let's see if we can help to provide such answers, instead of quarreling on unrelated issue.
Dear Marcel asked "why are chemical details not (always) provided on bottles or boxes making publicity, as is the case for food?" I dont think the answer is what you said there. It is very simple: such formula is a carefully guarded trade secret!
Dear mahmoud, psychology is not a science, you failed to provide evidence to the scientificity of psychology.
My answer to the original question remains:
Until you come out with a theory, testable and falsifiable. Your efforts are worthless
Dear Yaron. What you said originally was I quote you "Psychology, economics, sociology etc. are not sciences." I said "Yes. they are". I also provided evidence for one of those that you think belongs to fiction not science, that is economics. If I want to provide proof for other cases (psychology, sociology, biology, etc.) I must be an expert. But as I said before I am not. I am an engineer. So maybe if you want a good answer (better than I already provided) to your question you better ask experts in RG.
Aside from the attached paper (among others in my RG site) that may be used for transport and survival behavior of populations (wether be human or otherwise) I like to recommend the references in the paper such as 1, 6, 39 ..etc and especially 35 by Skellam.( seehttp://en.wikipedia.org/wiki/John_Gordon_Skellam) For it was Skellam That used brownian motion for the first time to describe Spread of muskrats in the German black Forests. Skellam is even credited with his own distribution ( see http://en.wikipedia.org/wiki/Skellam_distribution). In fact If I remeber correctly the Skellam 1951 paper was reprinted in another Ecological or Biological journal in the early ninetees, I believe. In fact if colleagues get a hiold of the PDF version of the paper I would appreciate a copy.
Kindest regards,
FBMB.
Salaam !
By the way I just found the Skellam paper: Here it is attached:
Kindest regards,
FBMB.
I found the Skellam paper interesting. Along similar lines, but more recent, I mention the paper: D. Brockmann, L. Hufnagel and T. Geisel, The Scaling Laws of Human Travel, Nature, 439 (2006), 462--.
Let me give us three parts. [1] Before responding to this question I wish to take, in this regard, a retrospective look at my own experience, say, in the year 1968. As a young assistant at Robert Reichardt’s Sociology Department, I was first and foremost confronted with (a) Herbert Simons “Models of Man” (1957) and (b) James Coleman’s “Introduction to Mathematical Sociology” (1964). Reichardt had a good feeling for the talents and carriers of his friends. Ten years later, Herbert Simon would obtain the Nobel Memorial Prize in Economics "for his pioneering research into the decision-making process within economic organizations" (1978). Coleman, in 1991, was elected President of the ASA. Reichardt and me tested, and often spoke about, a peculiar chapter in Simon’s book, namely “A Formal Theory of Interaction in Social Groups” (American Sociological Review, volume 17, April 1952). It represented an early attempt to model behavior in groups by some differential mappings. Clearly, the applicability was rather marginal. Coleman’s work developed further in terms of concepts of rational choice (Mathematics of Collective Action, 1973), that is, collective action had two layers - [namely (1) actors and (2) events and exchange determined by matrices of interest and control. Actors held control over events, events aroused interest of actors. Further, actors had measurable features of ‘power’ and events had measurable ’values’.] These things brought forth a dynamics of social exchange. – Today we know: not only the advantages of such models, but also their weaknesses and limitations. To mention just two of the deficiencies that became apparent over the last decades, these models of ‘man and action’ had no proper place for nonlinearity or what we call now ‘organization far from equilibrium’, no ‘risk’, and they had no place for ‘self-reference’. So let me, in the next point, give you an answer to your question which, I hope, is relevant, but perhaps disappointing:
You wish ‘to mathematically model the behavior of the humans’? Is it then not necessary to understand human behavior? Hence: do we understand human behavior? Let us begin with this simple observation: Our arts to design computers and systems of any kind, showing artificial intelligence, the evolution of artificial habitats and social cyberspaces, has it changed human behavior or not? I guess, our behavior has entirely changed. We need not give a list of all those changes. Hence, from this we can conclude that our skills of ‘mathematically modeling human behavior’ have the power to transform human behavior. Thus, behavior refers to itself via modeling. Any good model should be able to describe the transformation that it will bring about.
Investigating foundations, we run into serious troubles! Human behavior is fragmented and society has an increasing granulation. Humans behave like animals, but we do as if we act in accordance with unitary values. We do not stay tuned to our values! But the ‘world wars’ have shown that we skip those values every now and then. We seek harmony, equilibrium, peace, god, enlightenment, but in the end tear each other apart over oil, gas and money. To sum it up metaphorically, Marquis de Sade who said that men experience their ultimate lust in sexual fantasy and violence, was imprisoned in his century, but he became leading philosopher for our (post)modern times.
Further: Most of us are after some sort of salvation. We seek god or enlightenment. How will you model this? Most of us act out of fear. So we defend territories, we attack or take flight, just like animals do. But the cognitive dissonances that lead to our raids are very complex. They vary from individual to individual. The internet has increased the probability of emerging conflicts. There doesn’t seem to exist any unitary model for such chaotic behavior in our ‘risk society’. Last but not least, the model making mind, the mathematician who constructs the ‘model of human behavior’ is just a small part of the human being. Society is a whole, and the model building mind is just a small part of the whole mind. How can that small part design the whole? – There arises some small chance as soon as we look for invariants, the archetypal structures of cognitive orientation. For men have logic, possess psychological morphogenetic structures feelings and sentience, we have empathy and compassion, and probably, if some of these qualities are preserved under special conditions, we can make some predictions as to ‘what’s the count’. But men are living creatures and it takes a lot of nerve to say our models are alive. Models don’t feel compassion.
Dear Bernd Thank for the story and your own experience. I read them both with great interests to learn from your experiences. I already pointed that any attempt to model such behavior should be regarded as "approximation" So nobody claim that any model is 100% water proof. They are fist attempt but get more improve as they evolve. Let me say a similar story about "artificial" neural network (ANN) as you already mentioned artificial intelligence. The first attempt was in 1943 by McCulloch-Pitts (a neuroscientist, and a logician, respectively). Their mathematical was simple and far from what how biological neuron works. Later a book by Hebb called "The Organization of Behavior" provided a neuropsychological theory. He laid a foundation for a theory of thought.. While scientists have added a lot to the science of neurology, Hebb advanced the state of neurology so much all the way back in '49. In 60's Rosenblalt showed ANN in action by creating Perceptron. Since then ANN had friends and foes but In the year 1986, there was breakthrough in ANN research. Rumelhart,. Hinton and Williams proposed a learning algorithm for multilayered feed forward networks called Back propagation algorithm. Even this was a very important mathematical model but still can not full model how neurons in our brain work. I gave the story of how ANN evolved since 1943 till 1986. As time passes we get closer and closer to it. So that is the beauty of mathematical modeling. When Morteza asked this question I guess he was (or is) interested to know how "rational" behaviors of human can be modeled (approximated) and he did not know you are going to discuss fear, fight, war, terrorism,, etc. With due respect these are not rational behavior, are they?
There are a more studies, principals and methods. This problem is very complicated and one is very partly solved only. If you are beginning in this problem, I will recommend to start with neurons networks and fuzzy logic for very simply activity of human behaviors.
Please, let us be patient and go considerately. Mahmoud Omid gave an excellent reply, and I would like to give him an appropriate answer before the thread vanishes out of our field of vision!
Dear Mahmoud, thank you for your outstanding reply. [I will, in this thread, put everything between brackets that is connected with ‘non-rational action’] Indeed, where human behavior is concerned, it is most relevant to bring in the organization of neuronal networks and neuropsychology together with names McCulloch, Hebb and Rosenblatt. Many have heard these names, only a few know the name ‘Heinz von Foerster’. He was an early collaborator of McCulloch, and later with his Biological Computer Laboratory (BCL), a research lab of the Departments of Electrical Engineering, Uni Illinois, played a prominent role in this field. [Less of us know the dishonorable, political circumstances of his discharge from the University of Illinois.] I have recently put a work onto RG having name “Four forms make a Universe” to lay the foundations for the construction of a conscious neuronal net capable of the internalized space-time. It is interesting that, mathematically, we need three things to make a brain: (1) an identity, (2) two orthogonal signal forms (bit strings) and (3) one logic circuit that is known as logic XOR. Probably you have heard that it was proven that simple perceptrons are incapable of learning XOR, that is, linearly inseparable patterns. But exactly these elements may turn out fundamental for internal space in human,- as Arnold Trehub would say: retinoid space. [Now, ‘postmodern philosophers’ have shown how important the discourse of “fear, fight, war, terrorism,, etc.” actually is. Please, think about how (quantum) physics came upon. In order to understand these things one must study curricula, the life of, say, Kepler, and later, the lives of Bohr, Planck, Sommerfeld, Pauli, Heisenberg, Schrödinger, Einstein, Born, Oppenheimer, Weisskopf, and still later the Thirrings and still others. There were tremendous conflicts in the attempt to unify the anomalous Zeeman-effect with Bohr’s model, conflicts between Heisenberg and Pauli, Heisenberg and Schrödinger, heavy conflicts that first had to be consolidated! By – by whom or by what? (wave and particle) First by the authority of Bohr and Sommerfeld, later by the genius of Dirac who put quantum motion and spin under the roof of relativity. But these old masters always had the feeling that physics was still wrong. And they were right! It is still wrong and we got to improve it by first understanding action as both, rational and non-rational, and second understand quantum motion as self organizing system of phenomena which are both inner and outer and which we do not yet see clear enough. But we know they have to do with primordial existence of energy, or call it ‘entangled bit strings’, primordial space and emergence of time.(as for irrational action, see: Arthur I. Miller "137").
Dear Colleagues
Salaam!
Should you be intersted in Modeling and interconnections between modeling systems
may I suggest these two papers of mine also in this regard:
https://www.researchgate.net/publication/234114403_Triple_modeling_scheme_for_optimal_control_related_to_weighted_population_dispersal
and,
https://www.researchgate.net/publication/234114412_Modeling_the_stochastic_dynamics_of_Zakaat
We bekieve that while the first paper shows some of the interconnections between three different types of modeling schemes, with a lot more to be dug up, the second paper (according to my knowledge so far) is the first of its kind in terms of putting the problem of Zakaat (mandatory philantropy so to speak in the theological sense) in the arena of mathematical problems. While that paper is a at this point an orphan first step and may be a caricature of the real phenomenon, I am very well open to collaborations and extensions in this regard.
Kindest regards,
FBMB.
Article Triple modeling scheme for optimal control related to weight...
Article Modeling the stochastic dynamics of Zakaat
For Modeling Purposes of Human Behavior - Diret to the point.
Here you will find one page attached paper ( by Andrew Conway) statistically modeling linking human intelligence and "tourism" based on an "idiomatic" cultural "joke". I find you can hardly beat this in terms of brevity and succint automatic application of Statistics rules.
Dear Morteza,
By reading your question, the first idea that I have is the behavior of humans in financial markets and this research branch is called ; it tries to understand the psychological biases and to model the human behavior in order to understand the different puzzles that characterize the financial markets.
http://www.ipedr.com/vol12/103-N10006.pdf
http://www.researchersworld.com/vol3/issue3/vol3_issue3_1/Paper_03.pdf
As a mathematical modeller I'm often asked "why model something when we don't fully understand the process."
If you don't understand a process, then trying to model it is one of the many techniques which can be used to provide insight. A model can be used as an aid for identifying the questions or in helping to develop tools to measures phenomenon.
Mathematical models developed for behavioural research have been used to predict and control behaviour in applied settings, and they have guided research in other areas of psychology
These point comes from this paper by James E Mazur, see this link for more detail
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1472627/
As modellers we are doomed to failure, but as Samuel Beckett said, “Ever tried. Ever failed. No matter. Try again. Fail again. Fail better.” So at each iteration we aim to "fail better" in the hope of furthering our understanding.
Also ".. while the individual man is an insoluble puzzle, in the aggregate he becomes a mathematical certainty. You can, for example, never foretell what any one man will do, but you can say with precision what an average number will be up to. Individuals vary, but percentages remain constant. So says the statistician." - William Winwood Reade
Salaam!
Here is a link to an MAA review i wanted to share with my RGers colleagues pertinent to the question of modeling of social and human behvior:---Enjoy!!!
http://www.jstor.org/discover/10.4169/000298910X492871?uid=2134&uid=2&uid=70&uid=4&sid=21105371242393
http://www.jstor.org/discover/10.4169/000298910X492871?uid=2134&uid=2&uid=70&uid=4&sid=21105371242393
In fact, very valuable modeling book(s), I mentioned intrinsically in one of my answers when referring to one of my papers and have been mentioned in the MAA review I pointed to earlier, is Beltrami's contributions to this topic: Here is one of them directly downlable from science direct in the link:
http://www.sciencedirect.com/science/book/9780124046245
There are 2 popular theories in Psychology that models human behaviour:
1. Theory of planned behaviour.
2. Cognitive Dissonance theory.
In order to turn these theories into mathematical formulas, you may like to check my presentation: http://www.slideshare.net/NohaElprince/adaptivelab-talk
and the paper:
J. Ni, D. Kulic, and D.E. Davison, "A Model-Based Feedback-Control Approach to Behavior Modification Through Reward-Induced Attitude Change", American Control Conference, (Washington DC), 8 double-column pages, June 2013. Invited paper. Submitted Sept 2012.
Human cognitive behaviour can be subdivided into three levels:
1) knowledge-based behaviour that describes high-level problem solving;
2) rule-based behaviour that is determined by rules and behaviour learned in the past; and
3) skill-based behaviour that involves elementary human information processing and basic control tasks.
Skill-based behaviour can be measured by taking a cybernetic approach, in which the response paths of the human are described with control-theoretical models. Changes in the identified model parameters serve as an objective measure for adaptation of human behaviour.
http://www.kyb.tuebingen.mpg.de/research/dep/bu/cybernetics-approach-to-perception-and-action/human-control-behaviour-in-closed-loop-tasks.html
I just can't resist... 1. what they did is not what you told them to do; 2. what they said they did is not what they told you they did; 3. the way they remember it is not the way it actually happened; 4. in any group there's at least one person who will want the others to all get along; 5. in any group there's at least one person who will disagree with everyone else just to cause trouble; 6. in any group there's at least one person who doesn't care what everyone else is doing; 7. in any group there's at least one person who's entirely clueless as to what's going on; 8. in any group there's at least one person who believes in aliens and claims to have seen Sasquatch; 9. all medical research is tainted because some of the test subjects lie about the stuff they eat, the drugs they take, and the symptoms they have; 10. in any group of test subjects there's at least one person who has had or currently does have every symptom imaginable and knows someone who has had every disease known to man; and the list goes on. People are so interesting! Elvis told me that Sasquatch is actually an alien.
3 thumbs up for Dudley's message.
Mhich is why studying and modeling human behavior is so..tricky.
Large sample sizes for surveys or big data observations can wash some of those...human anomalies...out of the noise around the data.
But -fact- is Elvis told -me- that Sasquatch is not an alien - Dudley is! ; )
Dudley's message is entirely accurate, but says nothing about the possibility of modeling human behavior -- only that using the model is inherently difficult. The problem is that one has huge sensor noise (with lots of correlation effects) so, unless one can apply the Law of Large Numbers by working with huge samples, the uncertainty in the variables makes prediction/intervention often useless.
On a more serious note, I have worked with some human data, and have been exposed to a lot of medical data. (Many of my father's medical publications are available on RG.) One thing I have noticed is that human-related data seems to be much more clumpy than the instrument readings (power plant data) that I work with on a daily basis, which is often normally distributed, with some skewing and perhaps dual peaks occasionally. Fitting a normal distribution to clumpy data is more difficult. We expect to be able to characterize data by averages and standard deviations, but maybe we shouldn't expect this for some types of data, and human-related data in particular. I've done a good bit of Monte Carlo simulations of various non-human systems, using distributed variables. Maybe we should consider models with variables that are not distributed, but rather clumpy. There are so many aspects of human behavior that can be divided into 2, 4, or 8 categories, which is why there are endless books and seminars based on this approach. There are always outliers in behavior, but more often than not, people follow a limited number of patterns. For instance our family likes to joke about people behaving like one of: lion, beaver, otter, or golden retriever (the 4 Myers-Briggs categories in familiar vernaculars). I add skunk, weasel, sloth, and border collie, as everyone knows at least one of them. Why not use integers for some variables and real numbers for others in a Monte Carlo simulation and keep track of the two separately?
This is not an easy task specially if one intends to model all aspects of human behavior. Many of the works in the fields like history, social sciences, economy, management, philosophy, psychology, medicine, and many other academic fields are attempts to present a model for some specific aspect of human behavior. In fact, our knowledge about human beings are the result of this attempts. For instance, economic theories concerning supply and demand is based on assumptions about rational behavior of human beings which is actually a model for human behavior.
Human behaviors must be categorized before you can get your head clear on how to start modelling. So which human activity are we talking about? Speech, hearing, thinking, sleeping, eating, vision, ... what? To model each of these behaviors you will need to live many times. Researchers have been trying to model the vocal chords for example, over the past 200 years or more, the degree of complexity that has come to light is something no one could ever think of. One cannot just want to model a complex organism without isolating certain functions for a closer scrutiny to discover how it works and therefore how to replicate it. Modelling humans? Where do you want to start? Is it truly worthy of thought?
Still, no concise theory of how to model human behavior.
No suggested experiments to test these models except for psychobabble.
First of all , this question is a part of Applied Psychology , Biology and Criminal research. But in general, I think , to study over the modelling of human behavior we have to categorize the human population according to Economic and Cultural status first. Because, socio- economic factor plays a vital role to change from one kind of behavior to another kind of behavior. The class difference ( have-have nots) is the key factor to mixing / organizing the people for the formation of behavior, attitude and aim of life. Take for instance , the behavior of a Political man and a teacher will not be the same. Similarly , that of a farmer and a beggar. So let us talk about what kind of modelling we are going to study and for what purpose it will serve. Thank you all.
Can the Fuzzy Set Theory and the intuitionistic Fuzzy Logic not help?
Simulation is used when conducting experiments on a real system would be impossible or impractical: for example, because of the high cost of prototyping and testing, or because the fragility of the system will not support extensive tests, or because of the duration of the experiment in real time is impractical. How to apply simulation can be illustrated by looking at an example based on a bank’s customer service department. Let’s assume that we’re trying to determine the minimum staffing level needed to reach some established service level. We can develop a Quality of Service measure; average queue size customers cannot exceed N people. In order to solve this problem you need to have some basic knowledge about the system such as how many customers visit the bank, at what frequency, and how long it takes a teller to service an individual customer.
http://www.anylogic.com/use-of-simulation
Dear Prof. Vuyst, Your question is genuine. Fuzzy logic and Intuitionistic fuzzy logic can help to study the model. Before applying this theory it is necessary to specify the problem first. In our information world more than 75% information covers non random uncertainty that is fuzzy information. So it is natural the problem can be modeled by fuzzy logic. In Inventory management problem we have used Phi-coefficient test to identify the decision maker's attitude over order quantity. This article has been attached here. By this way we can develop a model over behavioral science exclusively.
Provided you have measures for 'interest' and 'measures' for control over outcome of exchanges, for 'interest in interests' and 'control over interests' and so forth, you can model rather complex systems of human collective action even by linear algebra or geometric algebra. But one has to be aware of the exact limitations. For example, if Jill has an apple, but is not interested in apples, she rather likes oranges, whereas Jack likes apples rather than oranges, but has an orange, it is likely that they exchange their fruits and attain an equilibrium of action. The mathematics is merely one of stochastic 2x2 matrices. Now you can ask, if there are conditions that could inhibit the exchange, and so on. Mathematical models of man usually lead into further considerations, and into never ending stories.
Never ending stories are very interesting when they are meaningful. So quality also has to enter the story and sometimes it gets the generic term ‘qualia’, which in some cases is handled as noise.
Here in Belgium it was prof. Etienne Kerre who worked out the Fuzzy Set Theory.
Yes, you mention that. The problem is that in Nibhana noise can be meaningful.
I recently published a book: http://www.deconstructconflict.com which breaks down identity-based conflict into 12 components and combines these elements into a mathematical equation which exposes the underlying causes and triggers of conflict and shows how different approaches to managing conflict can be employed effectively. The 'Conflict Equation' is an algebraic statement of how components of conflict and its management interact. It is not intended for quantification, but as a qualitative formula it is very effective at explaining what is going on and guiding behaviors around identity based conflict. Let me know if this is of interest to you - I'd be happy to speak with you about it further. (BTW - Identity-based conflict is the type of conflict that exists in religious, ethnic and political conflict. This same type of conflict is also common in family business - which is the core application in this book.)
- Badges
- Science method