Featured Machine Learning Projects
E-Commerce Clothing Classifier
Created a multi-class image classifier to automatically tag clothing items. Applied transfer learning techniques with PyTorch and evaluated performance using accuracy and confusion matrices.
Next Word Prediction with Transformers
Implemented multiple transformer architectures for next word prediction, including an encoder-only transformer from scratch, a GPT-style decoder, and fine-tuned GPT-2 models. Trained on Sherlock Holmes stories for text generation.
Loan Approval Prediction
Built a machine learning model to predict loan approval status based on applicant information. Analyzed key factors influencing loan decisions and implemented multiple ML algorithms to find the best-performing model.
Hybrid LSTM-LR Forecasting
Implemented a hybrid forecasting model for stock price prediction, combining LSTM neural networks and Linear Regression. The model analyzes historical stock data to predict future price movements with improved accuracy.
Receipt Classifier
A machine learning system that classifies receipts as either real or AI-generated using a hybrid image-text approach. Combines EfficientNet-B0 for image features and BERT for text features extracted via TrOCR.
SPAM Classification
A text classification model for detecting spam messages. Implements various NLP techniques and machine learning algorithms to accurately identify unwanted messages.