Saturday, September 04, 2004

What is Fundamental Modelling?

What is fundamental? Now that is a really fundamental question! I believe that any truth or concept for which one doesnt need a more fundamental truth or concept to explain can be termed as fundamental.In that sense, the whole body of physics laws and knowledge is non fundamental, but always tending or striving towards finding the fundamental.

The field of Economics bases itself on a reasoning and logic on human behaviour in its act ofutilizing and sharing scarce resources.This field is based on common logic and entirely not fundamental as for most of the laws and theories proposed by economists we can always ask the question why and why not?

Then if physics and economics are not fundamental then what is?Isnt there any thing which is fundamental?

There is.Things created by humans. For instance rules of a chess game is a fundamental concept as it is meaningless to ask why are the rules of the chess game as they are? The 'why' question doesnt arise as we arent explaning an observation, but creating a base concept totally brand new.So any realm of the world which originated brand new like the 'Chess' game or the 'C' Language doesnt need explanation as they are bound and manmade concepts. They are fundamental concepts.

Then what about mathematics?Can we treat the whole field of mathematics as pure and fundamental and manmade. The fact that the concepts of mathematics are used for explaining or creating non fundamental concepts of physics, biology and economics is secondary as the pure concept of mathematics doesnt stand on the physical world but the other way round. But then though fundamental, there are many questions in the field of mathematics which are unsolved. For instance the value of Pi is untenable and a disgrace to mathematical world. The concept of geometry is extremely intriguing as that definitely has originated from the relation with the physical world and sometimes I feel may not be as pure as numbers. The concept of numbers too owes its existence to reality.But though mathematics of numbers draws inspiration from reality, it doesnt try to explain anything physical or god made. It is purely abstract or rather the question we need to ask is would we have evolved a different kind of mathematics if we have had a different universe governed by different physical laws? Would there be a place where the concept of counting thought of as meaningless? If so what mathematical form would it take? With the imagination I have I cannot imagine a different kind of mathematics, though I can do so for other bodies of knowledge. Mathematics is as fundamental as one can get.

Fine. Now we know that there are lot of non fundamental theories floating around in the world and most of them are used to explain, predict and make decisions.Each theory is trying to model using mathematical and conceptual tools the system which is under scrutiny or of interest. Using such a model, one would be able to explain , predict and make decisions. Physics tries to model everything. Biology tries to model the living. Economics model financial behaviour. These models are no way fundamental. For instance let us take the stock value of Cisco. The value of it was
ever increasing during the internet boom era. So one of the reasons for the Boom one would say is the greediness of people to make a quick buck. So one factor which influences the stock value of Cisco is the greediness of people. So if we were to make a financial model to predict the future stock value of Cisco we need to also consider the greediness value of the people influencing the financial markets.

Is it feasible to list all such factors which will affect Cisco's stock?

In one way as everything is connected it is impossible as it would be a list of every event in the world. Let me give an example. Say a particular person (Frank) has a daughter (Anita) calls her friend's mobile, but due to the fact that there was slight problem in the telephone exchange or coverage she wasnt able to reach her friend. She is travelling in a train and a nice young man is seated besides her and she borrows his mobile to make the call. After the call she returns the mobile and strikes a conversation with him. She finds that he is a tourist to newyork and also finds that he is taking the wrong train. So she helps him out in finding his way and in that time they get attracted to each other and in next few months they agree to marry. The father is pleased and decides to arrange for a grand wedding and sells his stock to arrange for the grandeur. Yes he was holding Cisco's stock. His sale though a limited amount would definitely have have had an influence on the Cisco's stock value. So while modelling the Cisco's stock value should we include in our model, the faulty coverage of mobile service providers. It is infeasible to create such exact models as there would be millions of shareholders each one with a different story.

So is such a modelling meaningful?
May be not in the stock market, but may be somewhere else. May be.In certain industries such a modelling however crude is paramount in running the business. For instance the insurance industry definitely has to model risk however crude it might be, as but for the model how would they calculate their premiums and run a meaningful business.In case of predicting weather we definitely need to model the best we can, so that we can predict the future to avert catastrophies and loss of human life. These crude laws and models have helped us immensely though it hasnt been accurate and correct. We are better off with crude laws than without any. These non fundamental laws/models are refined every day and we are moving in the direction of fundamentality in all fields.

Modelling can be achieved through human logic, reasoning and experience or though brute force of pattern recognition or curve fitting. The latter doesnt have any preconceived notions and it operates merely on datapoints and ends with a statistical model which explains the present datapoints and predicts the future datapoints. The former introduces human reasoning and common sense also with the aid of the observed data points to come up with a feasible model to explain complex system behaviour.

0 comments: