Track
Problems
200 hands-on ML implementations
3D Geometry
Activation Functions
Classic ML
- Compute Entropy for a NodeEasy
- Compute Gini Impurity for a SplitMedium
- Compute Information Gain for a SplitMedium
- Decision Tree Best SplitHard
- Gaussian Naive BayesHard
- Implement Majority Class ClassifierEasy
- K-Means Assignment StepEasy
- K-Means Centroid UpdateMedium
- KNN Distance + Neighbor LookupMedium
- Naive Bayes Log-Likelihood (Bernoulli)Hard
- Random Forest Majority VoteMedium
- Ridge RegressionMedium
Computer Vision
- 2D Convolution (Image Filtering)Medium
- Anchor Box GenerationMedium
- Average Pooling 2DMedium
- Bilinear InterpolationMedium
- Color to GrayscaleEasy
- Gaussian Blur KernelMedium
- Histogram EqualizationMedium
- Image HistogramEasy
- Image Rotation (Nearest Neighbor)Medium
- Max Pooling 2DMedium
- Morphological Erosion and DilationMedium
- Non-Maximum SuppressionMedium
- ROI PoolingHard
- Sobel Edge DetectionMedium
Data Processing
Feature Engineering
- BinningEasy
- Cyclic EncodingEasy
- Frequency EncodingEasy
- Implement Min-Max NormalizationEasy
- Interaction FeaturesEasy
- Log TransformEasy
- Min-Max ScalingMedium
- One-Hot Encoding (Multi-class)Medium
- Ordinal EncodingEasy
- Polynomial FeaturesEasy
- Rank TransformEasy
- Robust ScalingMedium
- Streaming Min-Max NormalizationMedium
- Target EncodingEasy
- WinsorizationMedium
Linear Algebra
- Calculate Eigenvalues of a MatrixMedium
- Compute Covariance MatrixEasy
- Implement Cosine SimilarityEasy
- Implement Dot ProductEasy
- Implement Euclidean DistanceEasy
- Implement Manhattan DistanceEasy
- Implement Matrix NormalizationMedium
- Implement Positional Encoding (sin/cos)Medium
- Linear Regression Closed FormMedium
- Make Diagonal MatrixEasy
- Matrix InverseMedium
- Matrix TraceEasy
- Matrix TransposeEasy
- PCA ProjectionHard
Loss Functions
- Binary Focal LossMedium
- Cosine Embedding LossEasy
- Implement Contrastive Loss (Siamese)Medium
- Implement Cross-Entropy LossMedium
- Implement Dice LossMedium
- Implement Focal LossMedium
- Implement Hinge Loss (Binary SVM)Easy
- Implement Huber LossEasy
- Implement InfoNCE LossHard
- Implement KL DivergenceMedium
- Implement Triplet LossMedium
- Implement Wasserstein Critic LossEasy
- Label Smoothing LossMedium
- Mean Squared Error (MSE)Easy
Metrics & Evaluation
- Cohen's KappaEasy
- Compute Accuracy, Precision, Recall, F1Medium
- Compute AUC (Area Under ROC)Easy
- Compute Confusion Matrix with NormalizationHard
- Compute Mean Average Precision (mAP)Hard
- Compute ROC Curve from ScoresHard
- Compute Silhouette ScoreMedium
- Expected Calibration ErrorMedium
- Implement Micro-F1Easy
- Implement R² Score (Coefficient of Determination)Easy
- Intersection over Union (IoU)Easy
- Isotonic Regression CalibrationHard
- Log Loss (Per-Sample)Easy
MLOps
Neural Networks
- Batch Normalization (Forward)Medium
- He InitializationEasy
- Implement a Simple CNN Layer (NumPy)Medium
- Implement Dropout (Training Mode)Medium
- Implement Global Average PoolingMedium
- Linear Layer ForwardEasy
- Max Pooling ForwardMedium
- RNN Step Backward (Vanilla RNN)Medium
- RNN Step Forward (Tanh Cell)Easy
- Xavier InitializationEasy
NLP
Optimization
- AdaGrad OptimizerEasy
- Cosine Annealing LR SchedulerEasy
- Gradient Clipping (Global Norm)Medium
- Implement AdaDelta Update StepMedium
- Implement Adam Optimizer StepEasy
- Implement AdamW (Decoupled Weight Decay)Easy
- Implement Gradient Descent for a 1D QuadraticEasy
- Implement Nadam (Nesterov + Adam)Medium
- Implement Nesterov Momentum (NAG)Easy
- L-BFGS Two-Loop RecursionHard
- Learning Rate Scheduler (Linear Decay)Medium
- Logistic Regression Training LoopMedium
- RMSProp Optimizer (Single Update Step)Easy
- Warmup + Linear Decay LR ScheduleEasy
Probability and Statistics
- Bernoulli Probability Mass Function & MomentsEasy
- Binomial Probability Mass FunctionEasy
- Bootstrap Mean & Confidence IntervalMedium
- Chi-Square TestMedium
- Expected Value (Discrete Distribution)Easy
- Geometric Probability Mass Function & MeanEasy
- Mean, Median, ModeEasy
- One-Sample t-TestMedium
- Percentiles / QuantilesEasy
- Poisson Probability Mass Function & Cumulative Distribution FunctionMedium
- Sample Variance & Standard DeviationEasy
Recommender Systems
- Adjusted Cosine SimilarityMedium
- Baseline PredictorHard
- Catalog CoverageEasy
- Hit Rate at KEasy
- Item-Based CF PredictionMedium
- Jaccard SimilarityEasy
- Matrix Factorization SGD StepEasy
- Mean Rating ImputationMedium
- NDCG (Normalized Discounted Cumulative Gain)Medium
- Novelty ScoreMedium
- Popularity RankingEasy
- Precision and Recall at KEasy
- Rating NormalizationEasy
- Top-K RecommendationsEasy
- User-Based CF PredictionMedium
Reinforcement Learning
- Advantage ComputationEasy
- Discounted ReturnsEasy
- Generalized Advantage EstimationMedium
- Monte Carlo Policy EvaluationEasy
- One-Step TD Value UpdateEasy
- Policy Gradient LossMedium
- Prioritized Experience ReplayEasy
- Replay Buffer SampleEasy
- SARSA UpdateEasy
- Tabular Q-Learning (Single Update)Easy
- Value Iteration StepMedium
- ε-Greedy Action SelectionMedium
Time Series
- AutocorrelationMedium
- Cumulative ReturnsEasy
- DifferencingEasy
- Double Exponential SmoothingMedium
- Exponential Moving AverageEasy
- Lag FeaturesEasy
- Linear InterpolationMedium
- Moving MedianEasy
- Percent ChangeEasy
- Rolling Standard DeviationMedium
- Seasonal AverageMedium
- Simple Moving AverageEasy
- Weighted Moving AverageEasy
Transformers
No pages match that filter.