Home Loan Approval Prediction Model

This is a project that automates the home loan approval process using predictive classification models. I built 4 models: Logistic Regression, Decision Tree, KNN, and Random Forest. I then did model evaluation to determine the best model for the job based on the F1 score.

Effect of Geographical Merchandising on Sales

The project uses OLS regression as well as interaction between predictors to predict the effect of merchandising in various geographical regions on sales.

Handwritten Digit Classifier

The project involves output from handwritten digits classification. It contains two binary classifiers and one vector classifier using K-means Clustering algorithm in R studio.

Project SQL

This project involves creation of relational database tables as well as DDL, DML, TCL operations. As well as data analysis.

Sales Analysis

This Power BI project, involving data cleaning in power query, changing data types, adding hierarchies, merging tables, using DAX formulas, creating visualizations, and presenting the dashboard.

Emotion Detection Model

Developed a model to detect multiple emotions in tweets using NLP techniques, including Feed Forward Neural Networks, LSTM, and pretrained encoder models (BERT, RoBERTa, DistilBERT) etc. Evaluated model performance based on the F1 macro average score to balance precision and recall for multi-labels.