Frightera/Sample-Machine-Learning-Projects

Covers diverse deep learning architectures including LSTMs with word embeddings for NLP tasks, GANs for synthetic image generation, autoencoders for anomaly detection and dimensionality reduction, and CNNs for computer vision applications like COVID-19 detection and malaria classification. Projects span both TensorFlow/Keras and PyTorch implementations, with some incorporating transfer learning via pre-trained models (ResNet152, VGG19, BERT, TF Hub) and ensemble methods. Includes time-series forecasting with SARIMAX and Prophet, imbalanced classification techniques, and uncertainty quantification approaches across domains from healthcare to fraud detection.

No commits in the last 6 months.