- 🎓 Aspiring Data Scientist — Focused on building data-driven insights and AI models
- 🧠 Strong in Python, SQL, Machine Learning, Power BI, and Exploratory Data Analysis (EDA)
- 🤖 Building predictive models using LSTMs, Face Recognition, and Time Series Forecasting
- 🌍 Based in Kollam, Kerala
- 📫 Reach me at: LinkedIn Profile
End-to-end data science pipeline analysing electricity consumption & sales inequality across Indian states (2012–2021) — from raw government data to statistical hypothesis testing.
| Feature | Standard EDA | This Project |
|---|---|---|
| 📊 Inequality Metric | ❌ Basic stats | ✅ Gini Coefficient + Lorenz Curve |
| 🧹 Data Quality | ❌ Drop bad rows | ✅ 5 surgical anomaly fixes (LOCF, Interpolation, Epsilon) |
| 🧪 Validation | ❌ None | ✅ 4 Hypothesis Tests (ANOVA, T-tests, Pearson) |
| 🏛️ Data Source | Toy datasets | ✅ Official Ministry of Power datasets |
| 🗺️ Coverage | Single metric | ✅ 9 sectors × 28 states × 9 years |
Tech Stack: Python • Pandas • NumPy • Matplotlib • Seaborn • SciPy • Jupyter Notebook
A robust face recognition system designed to work in real-world conditions including masks, hats, and glasses.
| Feature | Basic Face Recog | My Face Recognition |
|---|---|---|
| 😷 Mask Detection | ❌ Fails | ✅ Works with Masks |
| 👓 Accessories | ❌ Struggles | ✅ Handles Hats/Glasses |
| ⚡ Speed | 🐢 Slow | ✅ Real-time Optimization |
| 📊 Accuracy | 📉 Low | ✅ High Precision |
End-to-end time series forecasting for Rossmann retail stores using Facebook Prophet.
| Feature | Traditional Methods | Prophet Pipeline |
|---|---|---|
| 🗓️ Seasonality | Manual Handling | ✅ Automatic Detection |
| 🏖️ Holidays | Complex to Add | ✅ One-click Holiday Effects |
| 📉 Trend Change | Linear only | ✅ Flexible Trend Detection |
| 🚀 Scaling | Hard to Scale | ✅ Production Ready |
| Category | Technologies |
|---|---|
| Languages | Python, SQL, HTML |
| Data Science | EDA, RFM Analysis, Customer Segmentation, Data Visualization, Inequality Analysis |
| AI / ML | LSTM, K-Means Clustering, Face Recognition, Time Series Forecasting |
| Statistics | Hypothesis Testing (ANOVA, T-tests, Pearson), Gini Coefficient, Lorenz Curves |
| Tools | Power BI, Jupyter Notebook, GitHub, VS Code |
| Libraries | Pandas, NumPy, Matplotlib, Seaborn, Scikit-learn, Prophet, SciPy, TensorFlow/Keras |
| Project | Description | Tech Stack | Link |
|---|---|---|---|
| Next-Word Prediction | Keyboard-style word prediction using LSTM models | Python, LSTM | View |
| Face Recognition | Real-time recognition with mask/hat/glasses support | Python, OpenCV | View |
| Project | Description | Tech Stack | Link |
|---|---|---|---|
| ⚡ Electricity Inequality | Gini + Lorenz analysis of India's power distribution (2012–2021) | Python, SciPy | View |
| Superstore Analysis | End-to-end EDA and visualization of retail data | Python, Pandas | View |
| Customer Segmentation | RFM analysis and K-Means clustering for retail | Python, K-Means | View |
| Sales Forecasting | Time series forecasting for Rossmann stores | Prophet, Python | View |
⭐ If you find my work helpful, please star my repositories! ⭐