Python & Machine Learning Projects

EDA - Amazon Products and Discounts 2023

Conducted Exploratory Data Analysis on a sizable dataset of Amazon products and discounts from 2023. Through data cleaning, uncovered patterns and trends and examined relationships between variables. These insights enable businesses to make data-driven decisions on the Amazon platform.

KKBox Music Recommendations

Created a Predictive model for KKBOX that accurately predicts users' repetitive song listening. Used a Decision Tree model for binary classification and evaluated performance with metrics like classification report, confusion matrix, accuracy and cross-validation scores.

Cardiovascular (Heart attack) Disease Prediction

Worldwide, cardiovascular illnesses are the primary cause of death. The goal of this study is to forecast cardiovascular disease by identifying the major risk factors for heart attacks and creating a reliable prediction system.

Mercedes Benz Greener Manufacturing Project

Mercedes-Benz Greener Manufacturing project aims to optimize the testing system for their premium cars. This is achieved through the reduction of testing time by implementing a robust algorithm capable of predicting testing durations based on various feature combinations.
Implemented label encoding, dimensionality reduction and used XGBoost for accurate predictions. This resulted in faster testing and lower carbon dioxide emissions while maintaining Mercedes-Benz's high standards.

311 NYC Customer Request Analysis

The goal of this project was to provide an in-depth analysis of the NYC311 service request data that was gathered in New York City. The project concentrated on improving abilities in Data Wrangling, Exploratory Analysis and Data Visualization by utilizing Data Science methodologies.
The objective of the investigation was to derive significant insights into the customer service trends and complaint patterns that are common in New York City's variable environment.

Bank Term Deposit Prediction

Built a Decision Tree classifier to predict whether a client will subscribe to a Term Deposit based on their demographic and behavioral data. This predictive model can help Banks tailor their marketing strategies for maximum efficiency.

Movielens Case Study

Conducted Exploratory Data Analysis on Movielens dataset and built predictive models using XGBoost, Random Forest and Decision Tree to forecast movie ratings. Engineered features, identified key factors influencing ratings and applied Data Analysis and Machine Learning to generate actionable recommendations.

Diabetes Prediction

Project aims to improve Healthcare diagnostic capabilities by creating a predictive model that can reliably identify a patient's risk of Diabetes based on particular diagnostic parameters. The source of dataset is National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK)