A structured collection of internship deliverables spanning Python foundations, statistical analysis, machine learning, real-world data exploration, production-oriented model serving, and business intelligence.
This repository documents my work during an 8-week Machine Learning Engineer / Data Scientist internship at YuvaIntern. Each task is organized as an independent module with notebooks, source code, reports, visualizations, and supporting assets.
The portfolio progresses from core data-science concepts to practical, end-to-end projects: an electric-vehicle population study, an e-commerce churn-prediction system, and an Adidas US sales analytics dashboard.
- Built foundations in Python, statistics, data types, EDA, feature engineering, and reproducible ML workflows.
- Developed practical data-analysis projects using real-world datasets.
- Created an e-commerce churn prediction and retention engine with a leakage-safe Scikit-learn pipeline and LightGBM model.
- Built executive-ready business intelligence reporting for Adidas US sales using Power BI, Power Query, DAX, and Excel.
- Documented each task with reports, notebooks, project assets, and clear folder-level READMEs.
Machine-Learning-Engineer-Internship/
|
|-- Task-1-Data-Science-Fundamentals/
| |-- Python, statistics, data types, DS concepts, and best-practice notebooks
| `-- Practical linear-regression assignment and report
|
|-- Task-2-Hand-on Data implementation/
| `-- Hands-on data implementation exercises and documentation
|
|-- Task-3-Real-World-Dataset-Analysis/
| `-- Electric-vehicle population analysis, notebooks, visualizations, and report
|
|-- Task-4-Data-Science-Tool-Mastery-Project/
| `-- E-commerce churn prediction application, model assets, notebooks, and API/UI code
|
`-- Task-5-adidas-us-sales-analytics/
`-- Adidas sales data, Power BI dashboard, reports, and visual assets
| Task | Focus | Key Deliverable |
|---|---|---|
| 1 | Data science fundamentals | Modular Python and statistics notebooks plus linear-regression assignment |
| 2 | Hands-on data implementation | Practical data-manipulation and analysis exercises |
| 3 | Real-world dataset analysis | Electric-vehicle market analysis with feature engineering, statistical EDA, and strategic insights |
| 4 | Data science tool mastery | E-commerce churn-prediction and retention engine with LightGBM and FastAPI |
| 5 | Business analytics | Power BI dashboard for Adidas US sales, profitability, and omnichannel performance |
An exploratory and statistical analysis of electric-vehicle adoption using real-world Washington State data. The project covers data cleaning, feature engineering, BEV/PHEV comparison, market concentration analysis, and geographic adoption patterns.
A production-oriented classification project that identifies customers at risk of churn using behavioral, engagement, and cart-friction signals. It includes feature engineering, class-imbalance handling, a Scikit-learn preprocessing pipeline, LightGBM modeling, and a FastAPI-backed interactive interface.
An executive business-intelligence dashboard analyzing 9,648 US retail transactions. It explores revenue, operating profit, product performance, retail partners, regional trends, and omnichannel fulfillment economics through Power BI and Excel.
| Area | Tools |
|---|---|
| Programming & analysis | Python, Pandas, NumPy |
| Machine learning | Scikit-learn, LightGBM |
| Visualization | Matplotlib, Seaborn, Power BI |
| Application serving | FastAPI, Jinja2 |
| Data preparation | Excel, Power Query, DAX |
| Documentation | Jupyter Notebook, Markdown, Word reports |
Clone the repository:
git clone https://github.com/durgesh693/Machine-Learning-Engineer-Internship.git
cd Machine-Learning-Engineer-InternshipFor Python-based tasks, create an isolated environment and install the task-specific requirements where available:
python -m venv venv
# Windows
venv\Scripts\activate
# macOS/Linux
# source venv/bin/activate
pip install -r Task-4-Data-Science-Tool-Mastery-Project/requirements.txtOpen the relevant notebook in Jupyter or VS Code, then run the cells in order. For the churn-prediction application, follow the instructions in its task-level README before launching the API/UI.
- Translating business questions into data-science workflows
- Cleaning, validating, and exploring structured datasets
- Applying statistical reasoning and hypothesis testing
- Building and evaluating regression and classification models
- Avoiding data leakage through train/test isolation and pipelines
- Communicating insights through dashboards, reports, and visualizations
- Packaging machine-learning work for practical use
Durgesh Maurya
Machine Learning Engineer Intern | Data Science & Analytics Enthusiast
GitHub · LinkedIn
If you find this portfolio useful, consider starring the repository.