A Method for Controlling Extrapolation when Visualizing and Optimizing the Prediction Profiles of Statistical and Machine Learning Models

TL;DR

This paper introduces a new method to control extrapolation in prediction profilers, enhancing the reliability of model exploration and optimization by avoiding invalid extrapolated predictions.

Contribution

The paper presents a novel approach combining extrapolation control with genetic algorithms for constrained optimization in prediction profilers.

Findings

Extrapolation control prevents invalid predictions in high-dimensional models.

The method improves the reliability of optimization results in real-world examples.

Simulation studies show effective avoidance of extrapolated solutions.

Abstract

We present a novel method for controlling extrapolation in the prediction profiler in the JMP software. The prediction profiler is a graphical tool for exploring high dimensional prediction surfaces for statistical and machine learning models. The profiler contains interactive cross-sectional views, or profile traces, of the prediction surface of a model. Our method helps users avoid exploring predictions that should be considered extrapolation. It also performs optimization over a constrained factor region that avoids extrapolation using a genetic algorithm. In simulations and real world examples, we demonstrate how optimal factor settings without constraint in the profiler are frequently extrapolated, and how extrapolation control helps avoid these solutions with invalid factor settings that may not be useful to the user.