Interactive Tutorials: Machine Learning for Heritage Preservation

Welcome! This repository contains a series of interactive tutorials designed to help heritage professionals use machine learning techniques—specifically Factor Analysis and Feature Importance—to prioritize preservation decisions.

Introductory Technical Barrier. These tutorials are designed as a guided experience. While you don't need to be fluent in Python, a willingness to engage with script-based analysis, execute code cells, and follow built-in troubleshooting guides is necessary. These tutorials run in your web browser using Google Colab.

🚀 Getting Started

To use these tutorials, you need:

1. A Google account (to run Google Colab).
2. The sample dataset (heritage_data.csv).

Step 1: Get the Data

First, you need to download the sample dataset to your computer.

1. Click here to view heritage_data.csv on GitHub
2. Click the Download raw file button (the icon that looks like a tray with a downward arrow) on the right side of the screen.
3. Save the file to a known location on your computer (e.g., your Downloads folder or Desktop).

Note: You will need to upload this file to Google Colab each time you start a new notebook.

---

📚 The Tutorials

Click the links below to open each tutorial directly in Google Colab. We recommend going through them in order.

1. Data Preparation

Open in Google Colab

• What it does: Teaches you how to "clean" your data so computers can understand it.
• Key skills: Handling missing values, converting text categories (like "Brick" or "Stone") into numbers, and standardizing different units.
• Output: A clean dataset ready for analysis.

2. Simple Statistical Methods

Open in Google Colab

• What it does: Introduces basic statistical tools to explore your data before using complex models.
• Key skills: Correlation analysis (seeing which variables move together) and simple regression.
• When to use: When you have a small dataset (fewer than 50 buildings) or just a few variables.

3. Factor Analysis

Open in Google Colab

• What it does: Finds hidden patterns in your data by grouping related variables.
• Key skills: Identifying underlying drivers of deterioration (e.g., "Moisture Stress" vs. "Thermal Stress").
• Goal: Simplify complex data into meaningful themes.

4. Feature Importance

Open in Google Colab

• What it does: Ranks your variables to see which ones best predict a specific outcome (like Condition Rating).
• Key skills: Using machine learning models (Random Forest) to answer "Which factor matters most?"
• Goal: Prioritize what to monitor or fix first.

5. Visualization

Open in Google Colab

• What it does: Generates publication-quality charts and graphs.
• Key skills: Creating correlation heatmaps, boxplots, and scatter plots to communicate your findings.

---

🛠️ How to Use Google Colab

When you open a notebook link, follow these steps:

1. Connect: Click the "Connect" button in the top right corner to start the server.
2. Upload Data:
   • Click the Folder icon 📁 on the left sidebar.
   • Click the Upload icon (a paper with an up arrow) at the top of the sidebar.
   • Select the heritage_data.csv file you downloaded in Step 1.
3. Run Code:
   • You can run each "cell" of code by clicking the Play button ▶️ next to it.
   • Or, go to the top menu and select Runtime > Run all to run the entire analysis at once.

Tip: If you get an error saying FileNotFoundError: [Errno 2] No such file or directory: 'heritage_data.csv', it means you haven't uploaded the data file yet! Just follow the "Upload Data" steps above and try again.