Designing stimulating programming assignments for an algorithms course: a collection of exercises based on random graphs

  • Authors:
  • Michael Mitzenmacher

  • Affiliations:
  • Univ. of California, Berkeley, Berkeley,

  • Venue:
  • ACM SIGCSE Bulletin
  • Year:
  • 1996

Quantified Score

Hi-index 0.00

Visualization

Abstract

The field of random graphs contains many surprising and interesting results. Here we demonstrate how some of these results can be used to develop stimulating, open-ended exercises for courses in algorithms and data structures or graph theory. Specifically, we provide problems for algorithms that compute minimum spanning trees, connected components, maximum flows, and all-pairs shortest paths.