Bachelor Thesis: Battleship App

This thesis revolves around creating a web application, which allows a user to play against a learning battleship engine, which improves after each game.

The project consists of two central parts.

The first half of the project involves creating a game architecture, where a user is able to play against the engine. It should allow all legal moves of the game and some quality-of-life features, which make the game progress more understandable. Past moves, hit ships, missed shots, remaining ships and others should be easily visible.

The second part of the thesis centers around developing a battleship engine that will compete with a user. After each game, the engine will try to comprehend the game moves and adapt its strategy. The engine will be based on a reinforcement learning technique, which allows it to improve and learn from past games results.

Creating Battleship with an AI engine from scratch presents a challenge with many sides. The game contains rules and logic, which the webapp should properly display and the AI needs to understand. The main difficulty lies in writing an algorithm, which is capable of learning from past moves and adapting accordingly. Not only retrospectively, but also prospectively.

Artificial intelligence is a widely discussed topic of conversation. Many are interested in the capabilities of AI. From a student perspective, creating your own AI algorithm is a captivating project, serving as proof of an undertaking befitting a bachelor thesis.

Number of Students: 1
Supervisor: Philipp Zech