Vatsal Bafna

I am currently pursuing my MSc in Applied Statistics & Analytics at NMIMS, Mumbai, where I have honed my skills in data analysis, machine learning, and Python programming. My academic background has equipped me with a solid understanding of statistical methodologies and their practical applications in real-world scenarios. I have also completed several projects. These projects demonstrate my ability to apply theoretical knowledge to solve complex problems and develop innovative solutions.

• Amazon Web Services: Sales Accreditation (05/2024)
• Amazon Web Services: Technical Accredited (05/2024)
• Udemy : The Ultimate MySQL Bootcamp , Go from SQL Beginner to Expert

Predictive Risk Analysis for Loan Default Using Machine Learning Models:

• To predict loan default risk using machine learning models like Logistic Regression, Random Forest, and XGBoost.
• After data preprocessing (handling outliers, encoding, and feature engineering), the models were evaluated using metrics like accuracy and AUC-ROC. XGBoost performed best with 91.2% accuracy and 0.92 AUC-ROC.
• XGBoost proved to be the most effective model, helping financial institutions improve risk assessment and decision-making.

Intermittent Demand Forecasting :

• Used 3 years of pharmacy data for 6 different medications to conduct a thorough investigation of
intermittent demand forecasting.
• Classified drug demand into Smooth, Lumpy, Erratic, and Intermittent categories using ADI and CV2.
• Applied different forecasting methods, including Croston and ARIMA, for performance evaluation.
• Employed evaluation metrics such as RRMSE and 14-day forecast error to identify most effective
model.

Stochastic Model for a visit to the Doctor’s Office:

• Implemented Semi-Markov Process to analyze a patient’s waiting time in hospital.
• After satisfying the model assumptions, the sojourn probability distribution was
concluded as Gamma distribution by using graphical and theoretical methods.
• Handled data processing and analysis using RStudio and MS Excel.

From Pixels to Paragraphs : Converting Video Files to Text Using Python:

• To create a UI that converts Video File into text format and provide a summary.
• There are two approaches used: Using an API, Python libraries and Stream lit.
• API provided 95% accuracy while using python libraries provided 75% accuracy.

• Maximum Number of Sales- Think & Learn Pvt. Ltd BYJUS June - Dec 2021
• Sales Ninja- Think & Learn Pvt. Ltd BYJUS Jan- June 2021
• 1st Prize in Inter-college Debate Competition (2017)