FINANCE PROJECTS
Stock Price Sentiment Analysis using News Headlines
Engineered a Word2Vec model to meticulously extract word embedding features from diverse news headlines. Leveraged a Random Forest classifier for precise prediction of stock price movements across multiple companies, addressing the challenging task of gauging market dynamics. Demonstrated exceptional model performance, attaining a remarkable accuracy rate of 85%, underscoring the model's efficacy in forecasting stock price fluctuations and aiding investment decisions.
Credit Card Risk Assessment using XGBoost
Designed and fine-tuned an XGBoost predictive model to assess borrowers' creditworthiness, effectively distinguishing between loan repayment and default scenarios. Employed rigorous hyper-parameter optimization through RandomizedSearchCV to identify and implement the most effective model configurations. Achieved an accuracy of 82.14%, highlighting the model's effectiveness in risk assessment and decision-making processes.
Stock Price Prediction using Stacked LSTM
Developed a sophisticated stacked Long Short Term Memory (LSTM) model to forecast the stock price dynamics of a prominent company. Employed a 100-day training window for each output prediction, ensuring the model's capacity to capture nuanced market trends. Additionally, created an engaging and interactive price graph showcasing the final predicted values, providing a comprehensive visualization of the model's forecasting capabilities.