Collaborative Filtering Movie Recommender
SVD-based matrix factorization system with 0.89 RMSE on MovieLens dataset. Handles sparse ratings, cold start problems, and scales to 100k+ users through optimized gradient descent.
SVD matrix factorization achieving 0.89 RMSE on MovieLens 100K dataset. Handles 94% sparsity through stochastic gradient descent with bias modeling and regularization.
Training Dynamics
SGD Optimization: Loss converges rapidly in 20 epochs. Optimal learning rate α=0.01, regularization λ=0.1 prevents overfitting to sparse patterns.
User Behavior Patterns: Rating distribution skews positive. Cold start performance improves rapidly with just 5 initial ratings.
Latent Factor Analysis: K=50 factors optimal — fewer underfit complex preferences, more overfit to training sparsity. Training time scales linearly with dimensionality.
Related projects
Entropy Wordle Solver
Information-theoretic greedy solver that picks each guess to maximize expected entropy over the remaining word set — averaging 3.92 guesses across 300+ games.
Decision Trees & Ensemble Methods
From-scratch implementation of decision trees with pruning, random forests, and AdaBoost. Comprehensive analysis of overfitting, feature selection, and ensemble performance on real datasets.
Neural Network from Scratch
Pure NumPy implementation achieving 99.6% MNIST accuracy through optimized gradient descent, backpropagation, and regularization techniques.