Predicting Insurance Premiums | EDA & LightGBM | Kaggle - Playground Prediction Competition (S4E12)

Developed a LightGBM model to predict insurance premium (1.2M records, RMSLE ~1.06), addressing challenges such as missing values, outliers, & weak direct feature correlations. For numerical features, created quantile-based binning and performed target-based statistical aggregation within each bin.

Predict Podcast Listening Time | CatBoost

Developed a CatBoost model for predicting podcast listening time (RMSE:12.83) using 5-fold KFold CV. Focused on a comprehensive Feature Engineering pipeline: Target-based statistical aggregation for selected variables. Listen ratio aggregation to capture content consumption efficiency.

Price Prediction | EDA | LightGBM | Kaggle Backpack Prediction Challenge (S5E2)

Developed a predictive model for backpack price estimation using LightGBM, achieving an RMSE of 38.93, placing within a very close margin of the top leaderboard score of 38.61.

Waze User Churn Prediction | EDA | Logistic Regression | Random Forest | XGBoost

Developed a user churn prediction model for Waze, covering EDA, Hypothesis Testing, Logistic Regression, and Machine Learning (Random Forest & XGBoost). Used Logistic Regression as a baseline and advanced tree-based models for better predictive accuracy.

Data-Driven Exploration of Renowned Artworks using SQL

A Comprehensive SQL Analysis of Masterpieces from Renowned Museums Across the Globe. This project aims to answer a series of 22 questions by leveraging the power of SQL queries to unveil profound insights and patterns, providing a nuanced perspective within the realm of art.

Minnesota Interstate Traffic Volume Analysis

Conducted an in-depth analysis of traffic flow on the interstate highway system, this presentation emphasizes four crucial elements: monthly volumes, weather impacts, peak-hour dynamics, and the influence of holiday travel.

Understanding Linear Regression: Building and Evaluating Simple and Multiple Linear Regression Models with Python

In this comprehensive guide, I delve deep into the world of Linear Regression, exploring key concepts, building both Simple and Multiple Linear Regression models using Python, and comparing the performance of these models to uncover whether adding more predictors truly enhances predictive power.

Sales Exploratory Data Analysis using Python

In this exploratory data analysis project, I delved deep into understanding the factors influencing sales generation, the distribution of sales across various industries, geographical sales patterns, and the product categories driving the highest sales volumes.