Christoph Emunds

Christoph Emunds

M.Sc. Artificial Intelligence, Software Engineer and Deep Learning enthusiast.

My Data Science Curriculum

2 minute read

I gathered a few links to video lectures and books about data science and machine/deep learning that I used or am using to get a deeper understanding of these topics.

  • [:heavy_check_mark:] Completed
  • [:eight_pointed_black_star:] In Progress
Python Data Science Handbook [:eight_pointed_black_star:]

This book is a good resource on how to work with data efficiently using numpy, pandas, matplotlib, and scikit-learn.

Mathematics for Machine Learning [:eight_pointed_black_star:]

I am currently working through this book to refresh my knowledge of certain areas of maths.

Andrew Ng’s Machine Learning Course [:heavy_check_mark:]

This is a more bottom-up approach on Machine Learning, starting from simple linear and logistic regression. In this course, Andrew Ng covers the basics of loss functions, regularization, and optimization. Eventually, he puts these concepts together into neural networks. He also has a section about Support Vector Machines, unsupervised learning, dimensionality reduction, anomaly detection, recommender systems and general advice on how to apply machine and deep learning in projects. The videos and reading material can be accessed for free.

Introduction to Machine Learning for Coders [:eight_pointed_black_star:]

In contrast to the course by Andrew Ng, this course by fast.ai is a more top-down approach to machine learning. Right in the first lecture the students are introduced to a hands-on problem involving the Blue Book for Bulldozers dataset using a Random Forest model. The course then drills down into the specifics of training different machine learning models.

deeplearning.ai [:eight_pointed_black_star:]

These video lectures are all part of the Deep Learning Specialization on Coursera:

  1. Neural Networks in Deep Learning
  2. Improving Deep Neural Networks: Hyperparameter Tuning, Regularization and Optimization
  3. Structuring Machine Learning Projects
  4. Convolutional Neural Networks
  5. Sequence Models
Practical Deep Learning for Coders [:heavy_check_mark:]

This course on Deep Learning by fast.ai consists of two parts. The first part is more like a practitioner’s guide, presenting an overview of best practices for problems in Computer Vision and NLP, but also giving some intuition about backpropagation and the training process. In the second part, Jeremy Howard goes the other way around. Starting from the foundations, he leads the students through implementing a lot of the fastai library themselves, explaining why things are done the way they are. Overall, the information presented in this whole course is very dense; so much that I’m watching it for the third time already and I’m still picking up new things.

CS231n [:heavy_check_mark:]

Stanford’s CS231n covers Deep Learning for Computer Vision applications. Together with the supplemental material and the assignments, this course helps you build a thorough understanding of the inner workings of neural networks in general and convolutional neural networks in particular.

CS224N [:eight_pointed_black_star:]

Stanford’s CS224N is a very thorough course for Natural Language Processing using Deep Learning. The lecture will give you a deep dive into the basics and provide you with the essential tools to easily understand the new advancements made in the NLP domain. The assignments for the current version of the course can be found on the course website.

The Deep Learning Book [:eight_pointed_black_star:]

I have yet to read this book cover to cover, but it can also serve as a reference to quickly look things up.

Categories:

Updated:

You may also enjoy