Subscribe to DSC Newsletter

Interesting thread on our LinkedIn group, with 15 replies so far.

Click here to read discussion.

Views: 96

Reply to This

Replies to This Discussion

Hmm, the same is in management consulting. "Best practices you recommended proved to be wrong and useless." managers say. "Why?" "Because we have too many reasons why we are not making use of them".

Do we really know what exactly our customer needs?
i dont't think so if assumptions are based on correct base .
Hi, Most of them in the beginning and few after an age. As per as my view & observation predictive models needs to be dynamic and should be trained,test & validate in frequent interval depending on the accuracy & senstivity of the model the age has to defined on the basis of domain & application.

No. All predictive models are models.

Models are not developed to reproduce reality -- they are there to effectively summarize experience along some relevant dimensions. The predictions say, in essence, that experience suggests that situations within the range of the model's dimensions should produce the indicated results within an error range. further, the indicated prediction is the best available given the data.

There are better statistical descriptions of what is going on (grab your graduate statistics text), but the above explanation usually works for clients.
Ms. Clarks statement has over time been taken somewhat out of context. In the original article she states that "Model users frequently forget that all models are based on simplifying assumptions, and therefore all models are wrong. Models attempt to replicate reality, but they are not reality".

In a sense she is right in that we should never presume that we have access to all variables involved in a model.

Here is the link to the original article.

-Ralph Winters
I think that, if you are talking statistics, you would have to say something about the randomnesss of the model's residuals.

If a model is to be statistically useful, that you are saying (among other things) that there are no systematic components in the residuals.
Models don't attempt to replicate reality. The purpose of model is to give simplified picture of certain relationships between certain aspects of reality (things, events, processes etc.).

The problem is how to make model simple enough to make it comprehensible for its Users and useful, and right enough to make it trustworthy.

In fact, this idea is absorbing me for years. Recently I even opened a blog on this - and related - matters. See


On Data Science Central

© 2021   TechTarget, Inc.   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service