Dynamic Programming Insights From Programming Contests

Competitive programming is a sport for solving highly complex algorithmic problems. These problems range from simple brute-force to advanced algorithms like network flows or dynamic programming. Many problems require the use of several different techniques at once. Practicing for these contests improves algorithmic problem solving, programming, testing, and debugging skills, and in general, makes students into well-rounded computer scientists. This thesis surveys over 70 problems from the Intercollegiate Programming Competition (ICPC) 2019 North American regional contests. These problems are fun and intellectually stimulating, and by solving them we have gained many insights into the technique of dynamic programming. These insights may be helpful in identifying and formulating dynamic programming solutions. This thesis can be used as an introductory text or supplemental reading material for the technique of dynamic programming.We built a web-based platform to archive ICPC problems and have archived all of the problems from the 2019 North American regional contests, totaling over 70 problems from 12 regional contests. While surveying these problems, we also classified them based on the algorithmic technique used. We hope that this website continues to archive ICPC problems from all geographic regions, and helps students identify problems for practice and training.