You are on page 1of 1

https://www.youtube.com/watch?

v=GIsg-ZUy0MY

In my opinion even if you go straight for deep learning you won’t go straight for
deep learning.

Go for the basics. The things you should cover by categories:

Math:

Stats: Everything you can. Stats are all around useful when you are building an
intelligent system even you are not using statistical model.
Calculus: From simple calculus you need to understand why are we calculating
derivative and why is it helpful for analyzing a function. In ML a method called
gradient descent is often used to minimize functions. If you want to go deeper into
calculus it will be helpful - for instance Lagrangian operator is used when
training support vector machines
Basic linear algebra: You need to understand all basic matrix operations and
why they are happening in such a way

Machine learning algorithms :

Start with the basic algorithms, learn the concepts behind them.

Nearest neghbour algorithms - super simple, but good to start with


Decision trees
Statistical - Naive Bayse for instance
Regressions - Linear, Logistic
Neural networks - play with multi-layer perceptrons, try different data,
different parameters
SVM
Something for clustering for instance kMeans

Feature extraction concepts:

Learn how to extract and work with features from the following domains

Text features
Image features
Time series

And finally there are other machine learning concepts you need to try out and get
comfortable with:

Evaluating systems: learn about different ways to evaluate your system - it’s
not only accuracy. Learn also about precision, recall, F1, MAP@N, AuC, etc.
Learn how to preprocess your data
Learn to visualize data, so you get better grasp on what’s happening
Ensemble learning
Dimensionality reduction
Hyperparameter tuning

After covering the basics you can start playing with deep learning techniques and
you will see that after knowing all this, it’s going to be much easier, then
starting on it now . Good luck !

P.s I will update this post if something comes to my mind, that I’ve missed

You might also like