Evolutionary Computation for the Automated Design of Puzzle Instances for Artificial Intelligence Education
Department
Computer Science
Major
Computer Science and Computer Engineering
Research Advisor
Tauritz, Daniel R.
Advisor's Department
Computer Science
Abstract
High-quality Artificial Intelligence (AI) education goes beyond traditional lectures by stimulating students’ desire to learn more deeply through Problem Based Learning (PBL). A typical approach in introductory AI courses involves providing the students’ challenging puzzles and games which capture the algorithmic complexity necessary to deal with real-world problem solving without overloading the students with having to deal with the messy details and scale of the real-world. However, it takes very significant effort to manually create such puzzles and games that are both intellectually stimulating and appropriate for the AI algorithms being taught. A critical component of puzzle design is creating a sequence of puzzle instances which differentiate the solving power of aforementioned AI algorithms. This project is concerned with automating the design of puzzle instances employing evolutionary computation in order to provide a high-quality hands-on learning experience.
Biography
Joseph Szatkowski is a senior pursuing degrees in Computer Science and Computer Engineering. He has previously taken classes in AI and Evolutionary Computing.
Research Category
Sciences
Presentation Type
Poster Presentation
Document Type
Poster
Location
Upper Atrium/Hall
Presentation Date
11 Apr 2017, 9:00 am - 11:45 am
Evolutionary Computation for the Automated Design of Puzzle Instances for Artificial Intelligence Education
Upper Atrium/Hall
High-quality Artificial Intelligence (AI) education goes beyond traditional lectures by stimulating students’ desire to learn more deeply through Problem Based Learning (PBL). A typical approach in introductory AI courses involves providing the students’ challenging puzzles and games which capture the algorithmic complexity necessary to deal with real-world problem solving without overloading the students with having to deal with the messy details and scale of the real-world. However, it takes very significant effort to manually create such puzzles and games that are both intellectually stimulating and appropriate for the AI algorithms being taught. A critical component of puzzle design is creating a sequence of puzzle instances which differentiate the solving power of aforementioned AI algorithms. This project is concerned with automating the design of puzzle instances employing evolutionary computation in order to provide a high-quality hands-on learning experience.
