Implementation of machine learning algorithms from scratch
The following algorithms/use-cases are implemented to date:
- Alternating Least Squares (ALS)
- Anomaly Detection using Autoencoders
- Understanding CNN Blocks - ResNet, Inception, Bottleneck etc.
- Decision Tree from scratch
- Denoising Autoencoder (DAE) on MNIST
- Entity Embeddings for categorical data
- Expectation-Minimization (EM) algorithm
- Fairness in ML
- Understanding Multi-Armed Bandits
- Multi-Task Learning with MNIST
- Entity Extraction with Named Entity Recognition (NER)
- Implementing Object Detection Metrics from scratch
- Understanding various OpenCV Transformations
- Semi-Supervised Learning
- Sequential Modelling with LSTMs
- Dealing with Sparsity
- Understanding Statistical Tests
- Thompson Sampling in Multi-Armed Bandits
- Time-Series Modelling with ARIMA
- Understanding Tokenizers & implementing Byte-Pair Encoding (BPE)
- Implementing UNet from scratch
- Image Similarity on MNIST using Contrastive Learning
- Factorization Machines from scratch
- Implementing RankNet - Learning To Rank (LTR) using Gradient Descent
- Probabilistic interpretation of AUC and MAUC (Multi-Class AUC)
Data Sources: Google