Can Mathematical models predict the economy?
Post Reply
 
Thread Rating:
  • 0 Votes - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
16-09-2013, 11:17 PM
RE: Can Mathematical models predict the economy?
I was a mechanical engineer by training and therefore somewhat versed in fluid dynamics. Until recently (about 15-20 years ago) some problems were so complex that hand calculations became too difficult and physical models had to be used to find out how specific systems worked. Now with the greater power of computers many of those physical models are not necessary and computer simulations may be used. This mind you with easily defined constraints, fluid viscosity, physical structure, temperature, pressure, etc.

I also noted on occasion that traffic patterns could be modeled using fluid dynamics as a starting point. It has been done and does provide some insight into traffic patterns but there is a huge problem, humans. How does one predict the frequency that 4 idiots will line up on a 4 lane freeway for 5 minutes at the same speed, less than that legally allowed, and prevent others from passing them such that a freeway gets jammed up (yes I have seen that).

I am afraid that those of you who seem to find fault with the assertion that mathematics cannot accurately predict how economics works are placing to much faith in getting all the variables in order, particularly the human one. I am not asserting that mathematics cannot be used to make some predictions on a micro level in economics. I do think that on a macro level mathematics starts running into the problem of too many variables.

How does one model even simple things like the desire of someone to own a home. I realized some 40 years ago that housing prices where I live (the San Francisco Bay Area) were in excess of what people with an average income in the area could afford. Despite that housing prices increased dramatically in the intervening years to the point that only people who's income very much exceeded the average could afford them, yet people continue to buy housing, how can that be.

Modeling human behavior such as economics is so complex that only simple systems can be successfully done. I think putting to much faith in mathematical modelling of economics on a macro scale is incorrect.
Find all posts by this user
Like Post Quote this message in a reply
16-09-2013, 11:27 PM
RE: Can Mathematical models predict the economy?
(16-09-2013 10:18 PM)frankksj Wrote:  There is something about human free will that science doesn't understand. I'm a software coder and have worked with artificial intelligence algorithms. For decades it had been speculated that, since the brain is composed of "simple" connections, once processing power reached a certain threshold, computers also would become self-aware and develop free will. Over the past decades the amount of raw processing power in computers is something they never could have imagined. But, we're no closer to having self-aware artificial intelligence. The best examples, like IBM's Watson, are still following set instructions so that given the same inputs, the output will always be the same. A lowly termite has free will, but Watson does not.

Thus, while I think mathematical models are useful and are at times good predictors, I do agree that until we understand what exactly makes up free will and can model it (which is at least decades away), we won't be able to develop any models that can be 100% accurate in predicting human behavior.

Chaos theory deals with systems that are highly sensitive to the initial conditions, so they seem random, however they are fully deterministic. Flipping a coin or rolling dice may seem random, but one could in theory calculate the result if one knew the exact initial position and momentum. The results of a coin or dice throw are in fact deterministic, but the precision of information one requires in order to calculate the result exactly is so high that the behavior approximates randomness. Markets can be modeled as "dynamical" systems where chaos theory can make useful predictions.

Free will may very well simply be an illusion. Our brains at a chemical and electrical level may act like a chaotic system does, with all the results of inputs being fully deterministic, but practically nearly impossible to calculate exactly. However chaotic systems are NOT random and are open to mathematical investigation.
Find all posts by this user
Like Post Quote this message in a reply
17-09-2013, 02:08 AM
RE: Can Mathematical models predict the economy?
(16-09-2013 08:56 PM)cjlr Wrote:  
(16-09-2013 08:14 PM)I and I Wrote:  The 90's mathematical predictions of the economy failed miserably for the same reason they all wouldn't work. Mathematical models to predict behavior of organic matter or living species is not possible. You can only do it on individual subjects but not a whole species or a whole solar system etc.

Are you implying that the mathematical equations to predict the economy in the 90's were correct?

I see you have simply reiterated your idiotic thesis. Some predictions are wrong, therefore all predictions are wrong. Keep reaching for that rainbow, champ.

Can you present a rule for determining which systems are amenable to mathematical models? There must be some sort of quantifiable difference, between such as are (cf: something you may have heard of), and such as are not (you provide "a whole species or a whole solar system", "organic matter or any behaviour of any living thing"). Really, now? Any behaviour of any living thing? A fascinating claim. How might one substantiate such a claim?

If you can determine which systems are open to mathematical description and which are not, please, publish your findings. There are a great number of people who would find such information utterly fascinating and extremely useful.

But let us leave that aside for now, and consider a related matter.

Nuclear physics. The rules governing nucleon interaction are well-known and well-attested. They are thorough and precise. The are verifiable through prediction (as they have been many times) and they are fully falsifiable (which they have never been). They do not permit an exact solution to describe the behaviour of complex atoms. And yet we may predict with a high degree of certainty how a radioactively decaying sample will behave; how electrons in bulk metals will behave; how solar wind and high atmosphere particles will behave.

Do you know how this is done? For the predictions are assuredly testable! And they have been shown reliable. What witchcraft is at work here? Well, I and I, I shall share the secret with you - long guarded by us delusional science sheep-believers, the secret to forming useful predictions with respect to complex systems.

Here.

Thumbsup


Science can only predict things when the variables of outside influence aren't counted.
Like your electron. Nice try though.
Find all posts by this user
Like Post Quote this message in a reply
17-09-2013, 02:15 AM
RE: Can Mathematical models predict the economy?
(17-09-2013 02:08 AM)I and I Wrote:  
(16-09-2013 08:56 PM)cjlr Wrote:  I see you have simply reiterated your idiotic thesis. Some predictions are wrong, therefore all predictions are wrong. Keep reaching for that rainbow, champ.

Can you present a rule for determining which systems are amenable to mathematical models? There must be some sort of quantifiable difference, between such as are (cf: something you may have heard of), and such as are not (you provide "a whole species or a whole solar system", "organic matter or any behaviour of any living thing"). Really, now? Any behaviour of any living thing? A fascinating claim. How might one substantiate such a claim?

If you can determine which systems are open to mathematical description and which are not, please, publish your findings. There are a great number of people who would find such information utterly fascinating and extremely useful.

But let us leave that aside for now, and consider a related matter.

Nuclear physics. The rules governing nucleon interaction are well-known and well-attested. They are thorough and precise. The are verifiable through prediction (as they have been many times) and they are fully falsifiable (which they have never been). They do not permit an exact solution to describe the behaviour of complex atoms. And yet we may predict with a high degree of certainty how a radioactively decaying sample will behave; how electrons in bulk metals will behave; how solar wind and high atmosphere particles will behave.

Do you know how this is done? For the predictions are assuredly testable! And they have been shown reliable. What witchcraft is at work here? Well, I and I, I shall share the secret with you - long guarded by us delusional science sheep-believers, the secret to forming useful predictions with respect to complex systems.

Here.

Thumbsup


Science can only predict things when the variables of outside influence aren't counted/or when in very controlled environments.
Like your electron. Nice try though.
Find all posts by this user
Like Post Quote this message in a reply
17-09-2013, 04:27 AM
RE: Can Mathematical models predict the economy?
(17-09-2013 02:08 AM)I and I Wrote:  
(16-09-2013 08:56 PM)cjlr Wrote:  I see you have simply reiterated your idiotic thesis. Some predictions are wrong, therefore all predictions are wrong. Keep reaching for that rainbow, champ.

Can you present a rule for determining which systems are amenable to mathematical models? There must be some sort of quantifiable difference, between such as are (cf: something you may have heard of), and such as are not (you provide "a whole species or a whole solar system", "organic matter or any behaviour of any living thing"). Really, now? Any behaviour of any living thing? A fascinating claim. How might one substantiate such a claim?

If you can determine which systems are open to mathematical description and which are not, please, publish your findings. There are a great number of people who would find such information utterly fascinating and extremely useful.

But let us leave that aside for now, and consider a related matter.

Nuclear physics. The rules governing nucleon interaction are well-known and well-attested. They are thorough and precise. The are verifiable through prediction (as they have been many times) and they are fully falsifiable (which they have never been). They do not permit an exact solution to describe the behaviour of complex atoms. And yet we may predict with a high degree of certainty how a radioactively decaying sample will behave; how electrons in bulk metals will behave; how solar wind and high atmosphere particles will behave.

Do you know how this is done? For the predictions are assuredly testable! And they have been shown reliable. What witchcraft is at work here? Well, I and I, I shall share the secret with you - long guarded by us delusional science sheep-believers, the secret to forming useful predictions with respect to complex systems.

Here.

Thumbsup


Science can only predict things when the variables of outside influence aren't counted.
Like your electron. Nice try though.

The influence of all possible variables which could impact an economic decision are accounted for, mathematically. You have no clue what you are talking about. But thanks for proving yet again, what am imbecile you are. You've obviously never even seen an econometric model.

Insufferable know-it-all.Einstein Certified Ancient Astronaut Theorist and Levitating Yogi, CAAT-LY.
Living daily with the high tragedy of being #2 on Laramie Hirsch's ignore list.
Find all posts by this user
Like Post Quote this message in a reply
17-09-2013, 06:29 AM
RE: Can Mathematical models predict the economy?
(17-09-2013 04:27 AM)Bucky Ball Wrote:  
(17-09-2013 02:08 AM)I and I Wrote:  Science can only predict things when the variables of outside influence aren't counted.
Like your electron. Nice try though.

The influence of all possible variables which could impact an economic decision are accounted for, mathematically. You have no clue what you are talking about. But thanks for proving yet again, what am imbecile you are. You've obviously never even seen an econometric model.

Accounting for every variable is not what we are talking about. We are talking about predicting (correctly) the variables that will take place. And no, mathematical models can't and didn't predict the current economic melt down.
Find all posts by this user
Like Post Quote this message in a reply
17-09-2013, 06:33 AM
RE: Can Mathematical models predict the economy?
(17-09-2013 06:29 AM)I and I Wrote:  
(17-09-2013 04:27 AM)Bucky Ball Wrote:  The influence of all possible variables which could impact an economic decision are accounted for, mathematically. You have no clue what you are talking about. But thanks for proving yet again, what am imbecile you are. You've obviously never even seen an econometric model.

Accounting for every variable is not what we are talking about. We are talking about predicting (correctly) the variables that will take place. And no, mathematical models can't and didn't predict the current economic melt down.


To clarify, are you saying that there are no current models that work or that no working model is possible?

Skepticism is not a position; it is an approach to claims.
Science is not a subject, but a method.
[Image: flagstiny%206.gif]
Visit this user's website Find all posts by this user
Like Post Quote this message in a reply
17-09-2013, 07:39 AM
RE: Can Mathematical models predict the economy?
(17-09-2013 06:33 AM)Chas Wrote:  
(17-09-2013 06:29 AM)I and I Wrote:  Accounting for every variable is not what we are talking about. We are talking about predicting (correctly) the variables that will take place. And no, mathematical models can't and didn't predict the current economic melt down.


To clarify, are you saying that there are no current models that work or that no working model is possible?

I am saying that there is no mathematical model that can predict accurately events that involve organic matter. The only models that exist that do work are models that take an example out of it's environment or study one sample of a species. Example: a tomato plant can easily be predicted but the future of tomato plants in 100 years cannot.

Humans are even much more difficult to predict as is the economy which is driven by human variables.
Find all posts by this user
Like Post Quote this message in a reply
14-10-2013, 07:44 PM (This post was last modified: 14-10-2013 07:56 PM by Bucky Ball.)
RE: Can Mathematical models predict the economy?
Seems the Nobel Committee thinks economic models can predict some things.
http://www.chicagotribune.com/business/b...9551.story

Insufferable know-it-all.Einstein Certified Ancient Astronaut Theorist and Levitating Yogi, CAAT-LY.
Living daily with the high tragedy of being #2 on Laramie Hirsch's ignore list.
Find all posts by this user
Like Post Quote this message in a reply
14-10-2013, 11:00 PM
RE: Can Mathematical models predict the economy?
A human is difficult to predict in the short term. Many humans together are relatively predictable. Even a single human is relatively predictable in the long term.

http://www.aaai.org/ocs/index.php/AAAI/A.../4845/5275

Give me your argument in the form of a published paper, and then we can start to talk.
Find all posts by this user
Like Post Quote this message in a reply
Post Reply
Forum Jump: