Credit Default Prediction System Using Machine Learning

• Built credit machine learning models with 150K rows to predict the probability of default payment • Implemented a “Imputation" strategy for the model using sklearn • Trained logistic regression, random forest and LightGBM models with an accuracy of 93.87%

Airline Satisfaction Prediction with Machine Learning

Cleaned & analyzed Airline Passenger Satisfaction dataset with 103904, rows & 23 columns for predicting passengers' satisfaction, Trained linear Models, Distance-based models, Tree-based models and Boosted tree using Light Boost using scikit learn.

Lemon-Free Rides: Predictive Modeling at Auctions

Worked with “Dont Get Kicked” dataset from Kaggle containing 70,000+ rows and 34 columns, Used machine learning models like logistic regression,random forest etc,Got the Best RMSE Score of 0.895 on the validation set after tuning.

Exploratory Data Analysis - Registered Companies in India (1900-2020)

Analyzed data for 1.9 million+ rows containing 17 columns by performing preprocessing and cleaning using Pandas, Created visualizations (scatter plots, pie charts, geo heatmaps & sunbursts etc.) using Seaborn & Plotly, Discovered maharashtra is the leading state in company registrations & nearly 60