The Fast and the Not-So-Frugal: Human Heuristics for Optimization Problem Solving

Genovefa KefalidouThe University of Nottingham, Nottingham, Nottinghamshire, United Kingdom
Tom OrmerodUniversity of Surrey

Abstract

In this paper, human heuristics have been identified that provide close to optimal solutions when solving Capacitated Vehicle Routing Problems (CVRPs). Results from previous experiments showed humans can produce good solutions relatively fast that compete with computer-based methods giving further support to previous research on Traveling Salesman Problems (TSPs). Multiple Regression analyses have been conducted to show the best heuristics adopted by participants and that lead to better CVRP solutions. Identified heuristics are categorized in visuospatial and arithmetic heuristics. Visuospatial heuristics (e.g. Clustering, Anchoring) performed better than the arithmetic (e.g. Balancing). Strategy switching appears to be a critical step within CVRP solutions suggesting that heuristics adopted are fast yet not-so-frugal, complimenting the fast and frugal toolkit. Results are discussed under the light of problem-solving theories and in terms of how best human heuristics can inform the current state-of-art computational algorithms used in optimization problem solving.

Files

The Fast and the Not-So-Frugal: Human Heuristics for Optimization Problem Solving (347 KB)



Back to Table of Contents