ANALYSIS OF US CONSUMER MORTGAGE COMPLAINTS:
Performed comprehensive analysis of U.S. consumer mortgage complaints, identifying borrower challenges in housing finance., Utilized data from Consumer Financial Protection Bureau and verified sources to analyze loan origination, servicing, foreclosure, and modifications., Employed advanced analytics to pinpoint high-risk regions and contribute to policy development, fostering equitable home finance industry.
SMART WATCH ANALYSIS:
Administered exploratory data analysis on Smartwatch Price Dataset to extract insights for a consumer electronics retailer., Leveraged predictive analytics and Python-driven data analysis to anticipate market trends and optimize product selection, inventory, and pricing strategies, enhancing decision-making and operational efficiency through cloud-based tools.
BIG TEETH REALITY TV:
Constructed a relational database for a reality TV show to manage detailed information on contestants, episode details, tasks, and voting records, ensuring the integration of sensitive health and task-related data., Expertly designed and set up a complex database structures to handle varied relationships like contestant details, background verifications, episode planning, and task distributions., Implemented key constraints to preserve record uniqueness and integrity, achieving a unique identifier system that maintained 100% accuracy in contestant email address exclusivity.
FAKE NEWS DETECTION USING NLP MODELS:
Collaborated on a Fake News Detection project, leveraging multiple machine learning models such as Logistic Regression, Naive Bayes, Random Forest, and BERT-tiny. Performed extensive data cleaning, profiling, and visualization to understand patterns in fake and real news articles. Implemented NLP techniques including tokenization, lemmatization, and TF-IDF vectorization to preprocess the data for model training. Achieved 99% accuracy with the BERT-tiny model, significantly outperforming traditional models in classifying news as fake or true.