Let's I have the a have a variable "A" which range from 80 (Min) to 50000 (Max) and the mean is 2500. I want to highlight both, low values and high values. Which function should I use to weight my variable and how can I do it. I have seen that sigmoidal/logistic function might work, but I am not familiar with it.
I know that the meaning of low values and high values must be strictly defined (I can use threshold based in standard deviation, or maybe something else).
The objective here is that I want to extract low values and high values and consider them as the characteristics of a particular phenomenon.
Thank you in advance!
PS
Dear Patrik,
I suggest you to see links in topic.
-Data Mining In Excel: Lecture Notes and Cases - CiteSeerX
citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.1003.1435&rep=rep1...
-Data Mining for Business Intelligence - X-Files
https://doc.lagout.org/.../Data%20Mining/Data%20Mining%20for%20Business%20Intell...
-Fitting Models to Biological Data using Linear and Nonlinear Regression...
www.mcb5068.wustl.edu/MCB/Lecturers/Baranski/Articles/RegressionBook.pdf
-Two Models of Population Growth - Wiley Online Library
onlinelibrary.wiley.com/doi/10.1525/aa.1980.82.4.../pdf
Best regards
When I familiarize myself with some of the definitions presented on the websites I can encourage you to use these functions (sigmoidal / logistic) and see the following sources for clarification:
1-https://stats.stackexchange.com/questions/167938/logistic-growth-models-for-count-data?rq=1
2-https://stats.stackexchange.com/questions/218905/logistic-regression-result-contradicts-intuitive-reading-of-visual-data?rq=1
3-https://stats.stackexchange.com/questions/204484/what-are-the-differences-between-logistic-function-and-sigmoid-function
4-http://www.biomath.bg/s%20markov%20publications/B2015-Kyurkchiev_Markov_Lambert.pdf
With wishes for success
Rana Z Al-Kawaz
It seems to be useful to adopt a triangular fuzzy formulation, which allows you to describe each variable with a triple containing the upper, the modal, and the lower value that the variable can assume
If you have a random variable with the minimal value A, maximal value B, and the most likely (or mean) value D and no other additional information, then the triangle distribution is the most appropriate one. It includes all boundary values A, B and D, i.e. T(A, D, B).
Let R is the uniform random number in the range 0
Sorry for my delay in respond your feedback's,
Thank you all,
Mohamed-Mourad Lafifi
Rana Z Al-Kawaz
Modestus Okwu
N. Epicoco
Alexander Kolker
I will give you further feedback as soon as possible.
You have helped me a lot.
Best regards,
Dear Alexander Kolker,
I really liked you comments, I have that idea in mind, but I have some difficulties to implement it, but now you made it so easy to me.
Thank you!
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts