Learning path
Data Science Portfolio Projects
Six end-to-end projects: sales analysis, churn, house prices, recommender, forecasting and an image classifier, with animated films and runnable Python. Part of the Data Scientist premium track.
-
1
Project 1: Sales Performance Analysis and Dashboard
A complete portfolio project: analyse a year of retail sales with pandas, find the top products, regions and trends, build KPI charts, and write the insights a manager…
-
2
Project 2: Customer Churn Prediction
Predict which telecom customers will leave: EDA, feature engineering, a scikit-learn pipeline, model comparison, threshold choice by business cost, and an explainable result.
-
3
Project 3: House Price Prediction
Predict flat prices from size, location and amenities: log-transformed targets, one-hot locations, linear vs random forest vs boosting, MAE in lakhs, and error analysis.
-
4
Project 4: Movie Recommendation System
Build "Because you watched…" recommendations three ways: popularity, content-based similarity on genres, and collaborative filtering with user-item ratings and cosine similarity.
-
5
Project 5: Retail Sales Forecasting
Forecast daily store sales for the next four weeks: calendar and lag features, festival effects, a gradient-boosting forecaster, a backtest against baselines, and a stock-planning recommendation.
-
6
Project 6: Image Classifier with Transfer Learning
Classify crop-leaf photos as healthy or diseased: dataset layout, augmentation, transfer learning with MobileNetV2 in Keras, evaluation with a confusion matrix, and a runnable small-scale version.