pattern-database

A highly modular Object-Oriented AI framework and interactive lab implemented in Python. Features automated benchmarking, pluggable solver engines, and interactive Pygame visualizers for Search, CSPs, Optimization, and Adversarial Games.

Interactive Rubik's Cube graph-search lab with COVE-3 solving, verified 3D replay, photo input, and a reproducible research paper.

🎓 Thesis: High-performance State-Space Search for Variant Slider Puzzles via Bitboard & PDB. | 本科毕业论文：基于位运算状态压缩与模式数据库的变体滑块谜题高性能搜索算法研究。

C++20 Rubik’s Cube solver with 1D, 3D, and bitboard cube models, BFS/DFS/IDDFS/IDA* search, and a corner pattern database heuristic

Optimal 8-puzzle solver: A* with disjoint additive pattern databases, benchmarked against Manhattan distance

An optimal Rubik's Cube solver that implements Korf's algorithm in C++.

Interactive N-Puzzle solver and performance analyzer utilizing advanced heuristics (Linear Conflict, PDBs) to empirically evaluate time and space complexities.

Finds every optimal Rubik's cube cross with an exact 190,080-state pattern database, then measures how the cross affects solve time across 750 of my own csTimer and smart-cube solves.