Code With Me : Decision Trees

Coding a decision tree from scratch!

[andyw]

pls do more of these coding videos! As a beginner I really learn a lot from your thinking and coding process

[adamcaudell]

We have been looking at decision trees under the hood for 3 weeks at Uni and you taught me more in 15 minutes. Thanks for the videos!!!

[taylorparrish]

More code with me pls! It's very helpful seeing you solve the problem in real time as well. Seeing your thought process really helps with the intuition.

[supakornsr]

This video really helps me understand the concept and the steps, although I am only using R. The length of the video and the amount of the contents are also perfect.

[sharmakartikeya]

Coding from scratch is the only thing I needed!
Thank you ritvik ☺️

[goitomyacobb.]

Simple & elegant. Kudos to you buddy.

[anuragsangem]

Thank you for this. I'll always learn something new in every single video of yours, please do more such videos.

[Uma]

Thank you... for making us to learn, think and code.

[tigert]

Thanks! This helps a lot! Really like your tutorials, because you make all this concept simple and easy to understand.

[process]

Awesome video! I didn't quite understand how to do the splits based on entropy but now I do. Thank you!

[chibuikejibuaku]

You're a good teacher. We should see more of this.

[anaydongre]

Great explanation.
Thanks for the coding session !!

[shubhampandilwar]

Great thinking process. I am trying to code some machine learning algorithms from scratch. This one helped immensely in my thought process. I am struggling to code logistic regression from scratch. Will you please help? Please do more of these how to code machine learning algorithms from scratch. As a beginner, it would help me and many others like me hugely.

[netanelmad]

Thanks! I love it. Maybe add some more explanations here and there for beginners :)

[BhuvaneshSrivastava]

Now this is really great !!
As always, really informative video as you actually went inside the code of decision tree.

Just a Suggestion:
It would be nice if you also show the python library (DecisionTreeClassifier) based implementation of decision tree and campare the results.
Also, is there a way for me to see the actual decision tree which gets created using decision tree classifier?

[wast]

How would you automatize the last part of updating the dataframe according to the best split and getting a new dataframe?

[chenqu]

So, after the 2nd split (using weight > 4), you used again the length feature for the 3rd split. Does that mean that the same feature can be reused in the nodes of decision trees? Would it lead to over-fitting ? Many thanks

[giorgikokaia]

Great explanation. Can you please share the csv file to run code?

[Phil-oymr]

Cool video! It may have been helpful to see how it runs with different steps sizes

I had trouble with the code when I used the fish column defined as 'type' I got all zeros, but no error message. Since type is reserved maybe?
So I had to change the column name to "fish_type" ...perhaps there is another way around this. Otherwise, it worked fine after that.