Projects

Project Overview

Project	Domain	Key Techniques	Data Type
Proj2	Real Estate (Buenos Aires)	Linear Regression, Feature Engineering	Tabular/CSV
Proj3	Air Quality (Nairobi)	Time Series, AR/ARMA Models	Time Series
Proj4	Earthquake Damage (Nepal)	Logistic Regression, Decision Trees	Classification
Proj5	Bankruptcy (Taiwan)	Random Forest, Gradient Boosting	Imbalanced Classification
Proj6	Consumer Finance (SCF)	K-Means Clustering, PCA	Clustering
Proj7	DS Lab Applicants	Chi-Square, A/B Testing, ETL	Hypothesis Testing
Proj8	Stock Volatility (MTN)	GARCH, API Integration, TDD	Time Series Forecasting

Lesson Workflows

Each project contains 4-5 lessons with predefined workflows. Example from Proj2:

Lesson	Topic	Model Pipeline
2.1	Price and Size	Size -> LinearRegression -> Price
2.2	Price and Location	Location -> Imputer + LinearRegression -> Price
2.3	Price and Neighborhood	Neighborhood -> OneHotEncoder + Ridge -> Price
2.4	Price and Everything	All Features Combined -> Full Pipeline -> Price

Data Sources

WQ/ ├── Proj2/data/ # Buenos Aires & Mexico City CSVs │ ├── buenos-aires--*.csv │ └── mexico-city--*.csv ├── Proj3/data/ # Nairobi air quality JSON │ └── nairobi.json ├── Proj4/data/ # Nepal earthquake SQLite ├── Proj5/data/ # Taiwan bankruptcy JSON ├── Proj6/data/ # SCF consumer finance CSV ├── Proj7/data/ # DS Lab applicants CSV └── Proj8/data/ # Stock data via API simulation

Cross-Project Analysis

The unified agent enables powerful cross-project queries:

# Apply technique from one project to another "Apply GARCH from project 8 to air quality data from project 3" # Compare models across domains "Which model works best for classification: Proj4 or Proj5?" # Educational integration "Explain how logistic regression works and show example from Proj4"