Satellite Imagery Geo-Analysis

Here's a paraphrase of the summaries for the tasks: **Extracting data from images (Task 1a):** * You'll need to add a new function to existing code that can grab specific values from images stored on Google Earth Engine (GEE). * These values, representing features of the land, will be used to train a machine learning model later. **Creating a new feature (Task 1b):** * You'll write some new code to define a new feature based on existing data (called NDVI). * This involves creating a new class that inherits from existing ones to handle this specific feature extraction. **Training machine learning models (Task 2a):** * Using all the features you extracted in Task 1, you'll train different machine learning models for a client. * The client has collected soil samples, and you can potentially combine their data with yours to improve the model's accuracy for their specific property. **Evaluating model accuracy (Task 2b):** * You'll need to assess how well the models you trained in Task 2a perform on the client's data. * This involves choosing appropriate metrics to measure accuracy and comparing the different approaches you used. * Based on the results, you'll decide which model performed best for the client's property. * This information will be crucial when communicating with the client in the next task.