🏠 Building a Regression Model from Scratch Using SEMMA (Python + Kaggle)
Ever wondered how real data scientists structure their projects?
In this video, I’ll walk you through how I built a house price prediction model from scratch using the SEMMa methodology — one of the most practical and process-driven frameworks in data science.
We’ll go step-by-step through each SEMMA phase:
💡 Sample – preparing and loading Kaggle’s Ames Housing dataset
📊 Explore – analyzing data distributions and correlations
🧹 Modify – handling missing data, encoding, and feature engineering
⚙️ Model – building and training a Linear Regression pipeline in Python
📈 Assess – evaluating results using MAE, RMSE, and R² metrics
By the end, you’ll understand how to turn messy real-world data into a working predictive model, and how the SEMMA framework keeps your workflow structured, clean, and efficient.