i'm an average engineering college student. i like machine learning and cinema. i sing sometimes.
developed a machine learning model to classify low RCS aerial targets of 6 different classes (e.g., drones, birds) using micro-doppler radar signatures. implemented attention-based CNN and LSTM models for improved feature extraction and sequence prediction. utilized ACGANs to augment a limited dataset. deployed the model on a web application, hosted via docker and AWS EC2, for real-time classification.
tech stack: python (pytorch, scikit-learn, matplotlib), flask, html, css, javascript, docker, aws ec2
model achieved an accuracy of 99.18% and an F1-score of 0.99.
deployed link: microclassify.streamlit.app
fetched 20 years of stock data using the yfinance library. implemented various algorithms and techniques such as LSTM for time series forecasting using libraries like tensorflow, scikit-learn, and keras. evaluated the model's performance using metrics like root mean square error (RMSE).
deployed link: stock-price-prediction-project.streamlit.app
developed both the front-end and back-end of the online chat app using socket programming in python.
created a custom search landing page for any browser that allows the user to access all important links in one click and search under any platform in the same search bar.
deployed link: startup-page-omega.vercel.app
ongoing project is twitter bot detection model.