All Projects

Panel Data Models - Applications in Macroeconometrics (Inflation Forecasting):-

Objectives: Evaluate and compare multiple panel data models (Pooled OLS, Fixed Effects, Random Effects, and Difference GMM) to forecast inflation across 77 countries (1980–2024).

Tools: Used Python (pandas, linearmodels, statsmodels), R (plm, AER), and SQL for data cleaning, imputation (MICE-BR), transformation, model estimation, and validation.

Strategies: 

• Constructed a balanced macroeconomic panel dataset (IMF WEO data)  

• Imputed missing values using MICE-BR at the country level  

• Conducted correlation and VIF analysis  

• Estimated and compared four panel models  

• Validated models using Hausman, Wald, and Sargan tests  

• Visualized results and created diagnostics

Results: The Two-Step Difference GMM model outperformed others in handling autocorrelation, heteroskedasticity, and endogeneity, making it the most robust model for forecasting inflation trends.

Human Resources Dataset Analysis:-

Objectives: Analyzing HR data to uncover factors influencing employee retention, satisfaction, and performance.

Tools: Used SQL and Python for cleaning, validation, querying, preprocessing, and model training.

Strategies: Database creation, exploratory analysis, feature engineering, and machine learning model development.

Results: Identified key retention factors and built accurate predictive models for employee attrition.

Systemic And Banking Crises:-

Objectives: Analyzing systemic and banking crises to understand their causes, impacts, and mitigation.

Tools: Used Python for cleaning, visualization, and data Preprocessing, for Analysis and modeling.

Strategies: statistical analysis, A/B test, econometric modeling, time series, and predictive Machine Learning models.

Results: Identified key crisis factors and developed accurate predictive models.

Loan Repayment Analysis: -

Objectives: Analyzed loan application data to assess borrower demographics and financial metrics for credit risk evaluation.

Tools: Utilized Python for data analysis and machine learning, along with libraries for statistical and predictive modeling.

Strategies: Conducted data cleaning and exploration analysis, and developed predictive models for loan default risk.

Results: Identified key factors influencing loan defaults and provided insights to enhance credit risk management strategies.

 Designing Bank (ATM) System and Database: -

Objectives: Designed and implemented a Bank (ATM) System and a relational database for managing banking operations.

Tools: Utilized Python for creating the Bank (ATM) System and SQL for database creation, querying, and management.

Results: Successfully streamlined banking operations with efficient data management and query performance.

Bike Share Data Analysis: -

Objectives: Analyzed bike share data to uncover usage patterns across cities and evaluated code runtime.

Tools: Utilized Python, NumPy, Pandas, and time for data cleaning, analysis, and visualization.

Strategies: Conducted exploratory data analysis, time series analysis, and user behavior segmentation.

Results: Provided actionable insights on user behavior, and peak usage times, and optimized Python code for better performance.