I have completed the first part of fast.ai machine learning course! I like that this course is opinionated. Jeremy Howards teaches about two machine learning techniques: random forests and neural networks. He is not trying to go through all the main machine learning algorithms superficially. Those two techniques, well mastered should be able to cover more than 80% of the industry needs. Here is the Pareto rule! I like that he is giving a lot of tips and tricks as well.
In this first post, I am sharing a list of points that I think are important to remember. This is a good exercise to go through these concepts quickly and be able to explain them with simple words.
Fundamental concepts
1️⃣ decision tree
2️⃣ random forest
3️⃣ rmse
4️⃣ training, validation, test set
Hands-on
1️⃣ take care of missing data
2️⃣ extract features from dates
3️⃣ one-hot encoding
Random forest interpretation methods
1️⃣ tree variance
2️⃣ feature importance
3️⃣ partial dependence
4️⃣ tree interpreter
Bits of software development
1️⃣ shell commands
2️⃣ object oriented programming in python
3️⃣ cython
This is my first post! I love my new blog. It is developed with hugo, it is so simple and lightweight! I like the design theme, it is minimal, readability is great, it helps to focus on the content!