When a new student is starting their journey into the field of computer science, it can be hard to figure out which topics to learn first. In this post, I will give my opinionated review on what every computer science student should know for their first internship. This advice is Read more…
Over the last few months, I’ve been working on a research project to develop personalized training tools for chess players. For this post, I’m just going to describe Maia, the framework developed last year which all of our code and models are based on. Maia Maia is a framework/model developed Read more…
Summary For this project, I analyzed methods to model the aleatoric and epistemic uncertainty of a neural network model for predicting the amount of time that a human player will use for a given move in a chess position. Dataset The dataset for these experiments comes from Lichess, an open-source Read more…
Summary For this project, I am comparing the abilities of Bayesian and non-Bayesian neural networks as a predictor for how much time a chess player will use to make their move in a certain chess position, taking into account the piece placement on the board as well as game metadata. Read more…
Problem During a chess game, players are constantly making decisions and evaluating the current state to increase their odds of winning. The process of analyzing positions is a problem that AI has already achieved superhuman performance on, and the use of these computer “engines” is essential for chess players who Read more…
Introduction In my AI class this semester, we have a 3-part course project based on the Berkeley CS118 Pacman Project. The first part of this project was due last week, and we were asked to implement some straightforward search algorithms like DFS and A*. However, the interesting part of this Read more…
In my last two posts, I gave brief overviews to two of the more established paradigms for chess engines, represented by Stockfish and AlphaZero. In this post, I will go over the latest engine on the block – Stockfish NNUE. NNUE was integrated into the main Stockfish branch in late Read more…
Last week, I wrote an article explaining the basics on traditional chess engines. Today, I will go over the new generation of chess engines, using neural networks instead of brute force evaluation. History For decades, neural networks were thought to be inferior to traditional brute-force engines because training a strong Read more…
Over the next few posts, I’ll be writing a series on how chess engines work. This isn’t intended to be an authoritative, in-depth guide on engines, but just a quick summary on the different types of chess engines, and how they compare against each other. History of Chess Engines Chess Read more…
Update – July 20, 2023 I’ve written two more blog posts on the topic of Colloquial Tamil. The first is an update to this post with a few additional learning resources I’ve found as well as published Anki decks. The second is a guide on how to read and pronounce Read more…