SANIKA DHAYABAR PATIL

Developed a high-accuracy fraud detection system using ensemble machine learning techniques, achieving 99.66% accuracy on test data.

Created an advanced image captioning system using deep learning and transformer models to generate accurate and contextually relevant descriptions.

Implemented probabilistic models for demand forecasting and inventory optimization, resulting in significant cost savings.

Developed an interactive learning platform for quantum computing concepts using OpenAI's GPT models to provide personalized explanations.

Built a compact and interactive guide on the fundamentals of causal inference, highlighting real-world use cases with Python-based examples.

Designed a high-performance computing pipeline to classify bone fractures from X-ray images using convolutional neural networks and parallel processing.

Developed a natural language processing pipeline for Hindi language tasks including tokenization, sentiment analysis, and text classification using deep learning models.

Implemented a neural network model to classify TMNIST textual image data, experimenting with multilayer perceptrons and optimizing performance through batch normalization and dropout.

Created an interactive math-solving chatbot that interprets natural language queries and solves algebra, calculus, and arithmetic problems using symbolic computation.

Developed a full-stack generative AI application that transforms text prompts into high-quality images using a fine-tuned Stable Diffusion model, integrated with CLIP, VAE, and U-Net components.