Using neural networks to solve grid-based puzzles

The gaming industry is expected to be worth over $100 billion in the next few years. From video games to mobile games; from board games to pen and paper games, there is a huge market for people with programming and mathematical skills to develop, design and refine games.

Dr. Herb Kunze works with his students to design neural networks that can be used to solve grid-based games such as Sudoku and KenKen. His students design the mathematical algorithm and the computer code to allow the computer to make decisions with partial memory to solve these puzzles.

This work is ongoing with both undergraduate and graduate students and Dr. Kunze is always looking for students to join his group and make valuable contributions.

If you have never tried a KenKen puzzle, learn about them and find one to try online! You'll be hooked!

Sudoku puzzle

High school students interested in developing the mathematical and statistical skills to tackle a real world problem such as this should consider our Mathematical Science Major, with an Area of Emphasis in Computer Science or Signal Processing.

Prospective graduate students interested in working with Dr. Kunze should visit his website, or read more about Graduate Studies at Guelph.