---
title: "A Learning System Which Accommodates Feature Interactions"
authors:
  - "Larry A. Rendell"
doi: ""
url: "https://ijcai.org/proceedings/1983/115"
pdf: "https://ijcai.org/proceedings/1983/0115.pdf"
venue: "Proceedings of the Eighth International Joint Conference on Artificial Intelligence"
year: 1983
pages: "469-471"
track: "Learning and Knowledge Acquisition"
keywords:
---

# A Learning System Which Accommodates Feature Interactions

Larry A. Rendell

*Proceedings of the Eighth International Joint Conference on Artificial Intelligence*, pages 469-471.

## Abstract

The author's state-space learning system has effectively optimized the coefficients of linear evaluation functions. The incremental approach uses statistical performance measures from completed solutions to bootstrap the heuristic, which estimates probability of task usefulness. These statistics are clustered in feature space, forming a mediating knowledge structure (region set) between the direct performance measures and the generalized evaluation function. The regions are data-determined, insensitive to noise, and allow management of interacting features through natural piecewise linearity. Early experiment with nonlinearity indicates stability, flexibility and improved task performance.