Statistical significance of a ML model...

  • Thread starter Thread starter fog37
  • Start date Start date
  • Tags Tags
    Linear regression
AI Thread Summary
To assess the statistical significance of machine learning models such as decision trees, SVMs, and neural networks, traditional tests like t-tests and F-tests used in linear and logistic regression may not apply directly. Instead, the discussion highlights the importance of uncertainty quantification (UQ), a developing subfield focused on evaluating model reliability and significance. A suggested approach involves setting aside a portion of the input data for testing, ensuring that the model is not evaluated on the data it was trained on, which would invalidate the results. For binary classification models, a scoring system can be implemented, where correct predictions are scored as 1 and incorrect as 0, allowing for comparisons against other predictive methods or random guessing to determine significance.
fog37
Messages
1,566
Reaction score
108
TL;DR Summary
Determining if a ML model is statistically significant...
Hello,

How do we check if a ML model is statistically significant? For models like linear regression, logistic regression, etc. there are tests (t-tests, F-tests, etc.) that will tell us if the model, trained on some dataset, is statistically significant or not.

But in the case of ML models, like decision trees, SVM, or neural nets, how do we determine if the model is statistically significant? I have not seen any specific test to do that...

Thank you!
 
Technology news on Phys.org
There is a whole subfield on this called UQ - uncertainty quantification. It is an area or active development.
 
fog37 said:
TL;DR Summary: Determining if a ML model is statistically significant...

But in the case of ML models, like decision trees, SVM, or neural nets, how do we determine if the model is statistically significant? I have not seen any specific test to do that...
The t test will work with any predictive model. You're supposed to set aside a part of the input data, and not use it in your model and use it for testing later. (Because predicting your input data with a ML model is cheating). For a yes/no model, you can score a 1 for correct, and 0 for wrong, and you can compare it other ways to predict the outcomes (or random guessing),
 
Thread 'Is this public key encryption?'
I've tried to intuit public key encryption but never quite managed. But this seems to wrap it up in a bow. This seems to be a very elegant way of transmitting a message publicly that only the sender and receiver can decipher. Is this how PKE works? No, it cant be. In the above case, the requester knows the target's "secret" key - because they have his ID, and therefore knows his birthdate.
I tried a web search "the loss of programming ", and found an article saying that all aspects of writing, developing, and testing software programs will one day all be handled through artificial intelligence. One must wonder then, who is responsible. WHO is responsible for any problems, bugs, deficiencies, or whatever malfunctions which the programs make their users endure? Things may work wrong however the "wrong" happens. AI needs to fix the problems for the users. Any way to...
Back
Top