Earth Data Science

This course covers the foundational aspects of data science, including data collection, cleaning, analysis, and visualization. Students will learn practical skills for working with real-world datasets oriented on earth data (seismic data, Lidar data, satellite images...)

Instructor: Antoine Lucas & Alexandre Fournier, Nobuaki Fuji, Grégory Sainton

Location: IPGP, 1 place Jussieu

Course Overview

This introductory course on machine learning covers fundamental concepts and algorithms in the field. By the end of this course, students will be able to:

  • Understand key machine learning paradigms and concepts
  • Implement basic machine learning algorithms using tensorflow and scikit-learn
  • Evaluate and compare model performance
  • Apply machine learning techniques to earth data

Prerequisites

  • Basic knowledge of linear algebra and calculus
  • Programming experience in Python
  • Probability and statistics fundamentals

Textbooks

  • Reference: “Machine learning avec Scikit-learn”, A. Geron
  • Reference: “Pattern Recognition and Machine Learning” by Christopher Bishop
  • Reference: “Deep leaning with Tensorflow” by A. Geron

Schedule

Week Date Topic Materials
1 Data for Earth science 1/2

Introduction to linear regression, other regression, and ACP.

2 Data for Earth science 2/2

Introduction to inference, inverse problem.

3 Crash course in deep learning (by Grégory Sainton)

Introduction to neural networks, gradient descent, backprop, regularisation…

4 Code Lab 1 - Inspect data and play with regressions

Learn how to clean improper or incomplete data. Apply several regression on them.

5 Code Lab 2 - Inverse problem on the Teil earthquake

Apply inverse problem to find the epicenter of the earthquake

6 Code Lab 3 - LiDAR data analysis and ML segmentation

Apply ML to analyse the multiscale data in LiDAR data

7 Assigment

Lab in group in 4h