Back to Study Plans

18 Problems

Build AlphaGo from Scratch

18 NumPy and PyTorch problems covering Go rules, a policy-value network, neural-guided MCTS, self-play, and training.

5 Easy8 Medium5 Hard

5 sections

Find a Go Group and Its Liberties Score a Terminal Position with Tromp-Taylor Rules Build the Legal Go Action Mask Apply a Go Action with Captures and Repetition Go Rules Quiz10 questions

Encode AlphaGo Zero Board History Apply an AlphaGo Residual Block Compute the Policy and Value Heads Assemble a Compact Dual-Head Network Policy-Value Network Quiz10 questions

Initialize Legal MCTS Edge Statistics Select an Action with PUCT Back Up an MCTS Simulation Path Neural-Guided MCTS Quiz10 questions

Mix Dirichlet Noise into Root Priors Turn Root Visits into a Search Policy Implement the AlphaGo Policy-Value Loss Self-Play and Training Targets Quiz10 questions

Sample a Recent Self-Play Replay Batch Apply a Dihedral Go Symmetry Apply One Policy-Value Head Training Step Run Deterministic Neural-Guided MCTS Integrated AlphaGo Zero Pipeline Quiz10 questions

Summary

Total18

Easy5

Medium8

Hard5

Progress0%

Quizzes0/5