How to implement Random Forest from scratch with Python

In the fifth lesson of the Machine Learning from Scratch course, we will learn how to implement Random Forests. Thanks to all the code we developed for Decision Trees, this implementation will be quite a bit shorter.

Welcome to the Machine Learning from Scratch course by AssemblyAI.
Thanks to libraries like Scikit-learn we can use most ML algorithms with a couple of lines of code. But knowing how these algorithms work inside is very important. Implementing them hands-on is a great way to achieve this.

And mostly, they are easier than you’d think to implement.

In this course, we will learn how to implement these 10 algorithms.
We will quickly go through how the algorithms work and then implement them in Python using the help of NumPy.

▬▬▬▬▬▬▬▬▬▬▬▬ CONNECT ▬▬▬▬▬▬▬▬▬▬▬▬

▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬

#MachineLearning #DeepLearning

[annawilson]

Great! To make it completely awesome, I guess n_features should be random as well, because in RF, the "random" aspect comes from two main sources:
-Each tree is built from a random subset of the data (known as bootstrap sampling).
-At each split in the tree, a random subset of features is considered.

[VritanshKamal]

Why didn't I find this playlist b4 !! Great content. !

[noura]

I've looked to DT and RF videos and they are very cool !!! By the way will you guys plan to upload video on gradient boosting?? Pleaaaseee ❤

[bzvn]

Hi, it's a good video, but I want to ask why you didn't implement the Random Subspace Method? Without it, it turns out that you have implemented bagging over trees. The Random Subspace Method is very important because it reduces error correlation between basic algorithms in random forest, which reduces variance of errors

[thomaswolff]

Excellent video! Could you add code for getting the out-of-bag accuracy metric from the random forest? Thank you!

[zelcadiana]

Hi. I am using random forest regression models to predict the mortality rate. My features have different dimensions, like millions, percents, thousands, etc. Do I need to do a standardization on my data before starting to built the models? Or any other kind of data transformation?

[mohamedhendy]

Can we use the same code for a regression task?

[pawlyk]

Great. Please, add the previous video to the playlist.

[exometria]

How do we print the predictions so we can see what it looks like? Just "print(predictions)?"

[sanpavlovich]

How about np.random.choice(n_samples, n_samples // 3)? It will correspond to random subsamples method and help to decrease correlation between trees, so it should improve accuracy. And thank you for video!