| Hi HN, I built a small AI-assisted tool that predicts which graph algorithm (e.g., Dijkstra, BFS, coloring) to apply based on graph properties. GitHub: https://github.com/lixiasky/Graphtune
Model: XGBoost trained on synthetic + real graph data
Use Case: Automatically selects the fastest algorithm for your input graph. --- But the reason I built it might be more interesting. A few days ago, I watched this video by a brilliant girl from Imperial College math department. She explained a Chinese math olympiad problem using Manim animations and really deep insights:
https://www.bilibili.com/video/BV1WsMCzeEQg/ Unfortunately, the video didn’t get much attention, and she mentioned on stream that she cried after seeing the low views.
So I made this project inspired by her problem and her visuals, hoping to help more people discover it. If you like this idea, you don’t need to star my repo — but please consider clicking into her video, giving it a thumbs up, and leaving a kind comment.
Maybe just a simple “nice work” will make a big difference. Thanks |