A Machine Learning Approach to Human Visual Learning

Principal Investigator: 
Project Overview
Background & Purpose: 

Because visual skills are often essential components of scientific expertise, it is important to discover to good ways of training novices to acquire these skills. We use ideas from the study of machine learning to develop new visual learning training techniques.

Setting: 

The University of Rochester, Rochester, New York

Research Design: 

The research design for this project is causal [experimental/statistical modeling]. This project collects original data using assessments of learning or achievement tests. The data are analyzed using standard statistical tests (e.g., ANOVAs) of differences in performances between the experimental and control groups. Computational models of learning to account for experimental data are also employed in this research project.

Findings: 

Our research suggests that information from other sensory modalities (e.g., audition, touch) can be used by the visual system for the purposes of visual learning.