ROSHAN KUMAR

M.Sc. Statistics & Computing graduate · Data Analyst

Varanasi, Uttar Pradesh, India

@roshan_kumar1

LinkedIn GitHub

0 followers

About

M.Sc. Statistics and Computing graduate from Banaras Hindu University with hands-on experience in Python, SQL, Power BI, Excel, machine learning, statistical analysis, and data visualization. Built end-to-end analytics solutions spanning ETL, SQLite data modeling, risk/performance analytics, forecasting, and interactive BI dashboards.

Experience

Data Analyst Intern

Bluestock Fintech

— – 2026

Mutual fund analytics capstone involving an end-to-end analytics pipeline for 40 mutual fund schemes, normalized SQLite star-schema database design, performance and risk analytics, and a four-page Power BI dashboard.

Education

M.Sc.

Banaras Hindu University

Statistics and Computing · 2026
B.Sc. (Hons)

Banaras Hindu University

Statistics · 2024

Skills

<3 months

R SQL NumPy Python Pandas Seaborn Matplotlib Regression Forecasting Statsmodels scikit-learn Dashboarding Data Cleaning Risk Analytics Feature Engineering Statistical Inference

Tools / apps / platforms

<3 months

Git GitHub SQLite Microsoft Excel Jupyter Notebook Microsoft Power BI

Projects

Retail Sales Performance Dashboard ↗

SQL, Power BI, Microsoft Excel

Built an interactive Power BI dashboard to track Total Sales, Profit, Orders, Customers, and Profit Margin using cleaned SQL and Excel data. Wrote analytical SQL queries and created regional, monthly, customer, category, and segment views with slicers and interactive filtering for business insights.
Air Quality Index (AQI) Prediction using Time Series Analysis ↗

Python, Pandas, Statsmodels, Prophet

Forecasted monthly AQI for Kolkata using 2017–2025 data after preprocessing, missing-value treatment, EDA, and time-series validation. Compared ARIMA, SARIMA, and Prophet using MAE, RMSE, and R2; Prophet achieved MAE 18.14, RMSE 21.84, and R2 0.83, with forecasts generated for Jan–Jun 2026.
Heart Disease Prediction using Machine Learning ↗

Python, Pandas, NumPy, Scikit-learn, Matplotlib, Seaborn

Built an end-to-end classification workflow on the UCI Heart Disease dataset with preprocessing, EDA, feature scaling, and model evaluation. Trained Logistic Regression, Decision Tree, and Random Forest models using stratified splitting, hyperparameter tuning, and cross-validation; compared Accuracy, Precision, Recall, F1-Score, ROC-AUC, and confusion matrices.

Courses & certifications

MS Excel · Simplilearn · 2025
Machine Learning · Great Learning · 2025
Data Science for Engineers · NPTEL · 2025
Data Analytics with Python · NPTEL · 2025

Explore the Broxer community

💬 Q&A Ask about roles, salaries and interviews — and help others. 👥 Circles Join discussions on skills, companies and industries. 📰 Feed Follow updates and insights from the community.