Optimization - Price Guidance 1.1.1

This document summarizes major improvements and fixes introduced in the Price Guidance Accelerator release version.

Version

1.1.1

Release Date

Table of contents:

New Features and Improvements

Description

ID

The feature analysis now recommends features until their cumulative Relative Importance reaches 90% instead of 80%, so more features can be recommended. New models preselect them in Select features on the Feature Preparation step. Existing models keep their saved selection, and only the Recommended column changes when you recalculate the feature analysis.

PFPCS-11238

Fixed Issues

Description

ID

For the Floor/Target/Stretch strategy, the Metrics tab of the Results step shows RMSE, MAE, MAPE and Bias, which do not apply to quantile models. Fix: the tab no longer shows these four metrics.

PFPCS-11247