a transport model (e.g logit model or ...) is based on statistical data and field works or merely based on mathematical theories or both of them?
If I want to define a model ( e.g. a new model in freight transportation ), what actions should I do? what kinds of data should I gather?
Obviously you need to define in detail which parts of the enormous system, that we may call "the logistics network." What do you want to capture with your model?
Also: before you start something new, do you know what parts of this enormous system you would like to provide a model for? You will definitely need to scale down from the global chains of transport and logistics - so what parts are you interested in? What is the purpose of this exercise? Could you, for example, start by reading the recent years of logistics networks papers, so that you can see what the models need, in terms of data, for example? Do you want to collaborate with a logistics partner, in order to obtain real data?
There are lots of questions to be answered, as what model you end up with will determine what questions will be possible to answer. This is not a five-minute exercise, but a PhD project, perhaps with actual partners from logistics.
Dear Sadat,
I am afraid my answer will disappoint you.
In general, and in transportation particularly, the academics consider varied type of model, according to what they want to achieve.
Maybe, the most important difference is between empirical model and theoretical models.
A theoretical model is only linked to a clear definition and the possibility to use math to reach some conclusion which can be qualitative. One of the best example is the “iceberg costs” see Samuelson : 1983 Thunen at two hundred. In this case the mathematical properties are very important.
A different case is when you have only data. See, for example Box Jenkins on Wikipedia.
In passenger’s transportation, we used to utilize a four steps model: from the generation to the distribution on the network. I am not sure it can be transposed to goods movement.
The very first question is : what do you want to do with your model.
a) Confirm a Hypothesis or find an exception to a law
b) Induce some new behavioural hypothesis
c) Estimate the future flows, in order to choose the capacity pf an infrastructure…
The list is longer than that.
As a scientist you have to follow a scientific approach considering you data and your empirical studies
- Badges
- Science topic
- Similar topics
- Mathematics