Skip to content

rohangawade/Predicting-Target-customers-for-Bank-Policy-subscribtion-using-Logistic_Regression_Transparency

Repository files navigation

Logistic_Regression_Transparency

CS-595 Interactive and Transparent Machine Learning

Predicting Target customers who will subscribe the new policy of the bank.

Finding the positive and negative evidences refering the Evidence-Based Uncertainty Sampling for Active Learning for Logistic Regression (https://link.springer.com/article/10.1007/s10618-016-0460-3) For the following types of objects, Find a) the total positive log-evidence, b) the total negative log-evidence, c) probability distribution, d) top 3 features values that contribute most to the positive evidence, e) top 3 feature values that contribute the most to the negative evidence. Print this information on the following object types

The most positive object with respect to the probabilities. The most negative object with respect to the probabilities. The object that has the largest positive evidence. The object that has the largest (in magnitude) negative evidence. The most uncertain object (the probabilities are closest to 0.5)

About

CS-595 Interactive and Transparent Machine Learning

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published