Explain information value in Machine learning

It is the process used by the industries for reducing the major variables before fitting the model as a number of variables present in the final model would be about 10. It is used in the basic filtering of the variable and decided whether to include those in the model or not. It works by the following formula. 

The following chart shows the predictive power of the model based on IV value.

In the following table, the price has been discretized based on price range and the counted number of events and non-events in that bin. The information value has been calculated for all the segments and added together for which we got the total value as 0.0356, which implies it is a weak predictor to classify events.