Machine Learning for Earth Sciences (AY: 2022)

This course is intended to provide a hands-on introduction to Machine Learning with a focus on its applications for researchers in Earth Sciences.

Prerequisites:

The following are assumed for each participant,

• Undergraduate level math proficiency.
• Acquainted with scientific research methodology.
• Intermediate level Python proficiency
• No background in ML.
• Basic knowledge of various domains within Earth Sciences.

Note:

• Python will be used as the programming language for instruction.
• The course proceedings will be hands-on and will require you to either use local python installation or Google Colab.
• Lecture content will not delve into the mathematical derivations of ML algorithms, but will cover the intuition and the necessary math to properly utilize the algorithms.
• This course will not have an assigned book as a primary reading reference, but will instead list out books, blogs, YouTube videos, GitHub repositories and such to serve as references that students are encouraged to explore.
• All necessary materials will be provided in the form of hand-outs, code exercises, recordings and such.

Lecture-wise breakdown:

No:	Title	Materials
1.	Getting Started with Machine Learning Introduction to machine learning and data science ML in real-life, academia, and earth science applications Conda environment setup ML Taxonomy End-to-end ML Pipeline	[Slides]
2.	Basic elements/building blocks under ML	[Slides]
3.	Basic elements/building blocks under ML (Continued...) Data Preparation Train-Validation-Test Overfitting and Underfitting Exploratory Data Analysis Brief introduction to Decision Tree and Linear Regression Ensemble Learning ML Taxonomy	[Slides]

References:

Coursera

• Machine learning Specialization by Andrew Ng
• Deep learning Specialization by deeplearning.ai
• Advanced Machine Learning Specialization by HSE University (Specialization under maintenance, so meanwhile use its GitHub repo)

GitHub

• 100-Days-of-ML-Code
• homemade-machine-learning
• zero-to-mastery-ml
• Machine learning Tutorials

Books

• Approaching (Almost) Any Machine Learning Problem by Abhishek Thakur (Book and Codes)
• Deep Learning by Ian Goodfellow and Yoshua Bengio and Aaron Courville
• Dive into Deep Learning

Getting Started with Python

• Python Programming And Numerical Methods: A Guide For Engineers And Scientists
• IPython Cookbook
• Python for Everybody Coursera Specialization by University of Michigan
• Python 3 Programming Coursera Specialization by University of Michigan
• Think Python: How to Think Like a Computer Scientist by Allen Downey

Some Good Jupyter-books or online books

• Data Science Topics
• Elements of Data Science
• Think Stats: Exploratory Data Analysis in Python
• Think Bayes: An introduction to Bayesian statistics using computational methods
• Computational and Inferential Thinking: The Foundations of Data Science
• Machine Learning from Scratch
• Interpretable Machine Learning
• Climate Change Impact Assessment: A practical walk-through
• Earth Data Analytics
• Computational Neuroscience

Some good competitive coding websites

• Hackerrank
• CodeChef
• Exercism
• CodeInGame
• HackerEarth