Machine Learning Diagnostic
Sample objectives, updates, model evaluation, representation, networks, and clustering.
Machine Learning
Classification Practice
Connect logistic probabilities and losses to threshold-dependent evaluation.
Convolution Practice
Compute convolution geometry and parameter counts accurately.
Dimensionality Reduction Practice
Relate PCA variance objectives to covariance eigenvectors and SVD.
Gradient Descent Practice
Compute deterministic and stochastic updates and diagnose step-size behavior.
Neural Networks Practice
Practice forward computation, backpropagation, activation behavior, and parameter counting.
Optimization Practice
Relate stationary points, convexity, objectives, regularization, and training diagnostics.
Regression Practice
Connect squared loss, gradients, normal equations, and ridge regularization.
Theory Practice
Reason about bias, variance, overfitting, generalization, and evaluation splits.
Unsupervised Learning Practice
Optimize and interpret hard clustering and mixture-model assignments.