Samarjeet Singh Chhabra

An experienced Data Scientist, worked as Artificial Intelligence and Machine Learning Intern at Tradelyzers. An AlmaBetter's Data Science program graduate with advanced courses and certifications from top platforms and universities. Skilled at solving complex problems, Thrive on diverse challenges across domains. The journey includes hands-on machine and deep learning projects, mastering predictive modeling and data insights. Dedicated and up-to-date, Ready to make a data-driven impact. Let's shape the future together.

Cardiovascular risk prediction 

AlmaBetter Verified Project -Supervised classification project

• Analyzed CHD dataset for insights and 10-year Coronary Heart Disease risk prediction. Conducted data wrangling, visualization, and feature engineering. Developed ML models (Naive Bayes, KNN, Logistic Regression, Decision Tree, SVC, Random Forest, XGBoost, Gradient Boosting). Achieved exceptional performance with F1 score 0.89, Recall 0.85, and accuracy 0.90 using Gradient Boosting. Enabled precise 10-year Coronary Heart Disease risk prediction for better healthcare planning and patient outcomes.

Customer Churn Prediction 

Deep Learning Web Project Deep learning web application

• Explored Customer Churn dataset to gain insights and develop model for predicting customer churn to improve customer retention and avoid customer losses. Conducted EDA and data visualization. Implemented feature engineering to enhance model performance. Explored various supervised algorithms on validation data and developed Neural Network for predicting customer churn. Established streamlined pipeline for data preprocessing, model training, and evaluation using OOPs concepts. Created a user-friendly web application for prediction model and deployed it on amazon web services EC2- http://51.20.9.125:8501/

FarmPet Robot- Major Project 

Bachelor's Major Project 

• Engineered a versatile, cable-driven agricultural robot for modern farming tasks, with full 3D mobility. Developed an innovative stepper motor control algorithm, leveraging a Microcontroller and Python scripting, to precisely regulate motor speeds and directions, optimizing agricultural processes. Will Enhance farming operations by increasing productivity and resource efficiency, leading to reduced labor costs and improved crop yield.

Bike Sharing Demand Prediction

AlmaBetter Verified Project- Supervised Machine Learning project 

• Explored Bike Sharing Demand dataset for booking factors and hourly predictions. Developed regression models for demand forecasting. Designed data pipeline, feature analysis, and enhancements for improved model performance. Identified key factors impacting bookings and Achieved 99% accuracy using XGBoost, optimizing operational efficiency. Optimized bike fleet management and resource allocation for improved operational efficiency, profitability and customer service.

USD-INR Timeseries Forecast 

Timeseries Forecasting

• Applying ML and DL techniques, we analyze and predict USD-INR conversion rates from Jan 2010 to May 2023. Performed data cleaning, model development, training, and evaluation. Gained trend of the data. Used Deep learning model architecture such as Multilayer Dense, Simple RNN, LSTM, Prophet neural model and SVR. Insights aid traders, investors, and policymakers in understanding and forecasting exchange rate dynamics.

Walmart Inc Sales Dashboard

PowerBI Dashboard 

• Analyzed Walmart Inc sales data across Myanmar cities using Power BI for strategic insights and profitability enhancement. Utilized SQL for efficient data extraction and Microsoft Excel for comprehensive data analysis and interpretation. Designed an interactive Power BI dashboard with filtering, drill-downs, buttons, and slicers. Incorporated decomposition trees, graphs, and charts to facilitate data-driven decision-making for sales improvement. Enhanced business decision-making and profitability.

Developed a predictive model for Standard Bank Africa, For loan approvals, leveraging customer data, yielding remarkable results with a notable 78% accuracy rate using Decision Tree.