@lerabyte: Random Forests Explained Most people learn decision trees as if they’re just flowcharts… but the real magic is in how they choose that very first question. Every split is a competition between impurity, thresholds, and how much cleaner each subgroup becomes. In this video, I’m walking you through exactly how a model evaluates different features (income, missed payments, total debt) and decides which one gives the strongest “signal” for predicting default. If you want to actually understand how machine learning makes decisions, this is one of the best places to start. Python coming in part 2! #python #ml #machinelearning #randomforest #rf
I was confused about how decision trees differ from random forests but this video explains it very well. Thank you!
2025-11-27 00:08:44
15
Tatiana :
Could a random forest also be used to predict whether a customer will make a purchase based on their past credit card behavior? What are your thoughts on that?
2025-11-27 03:35:12
7
Rayondemeil Studio :
Statistical prediction is well and good until something like COVID-19 arrives and completely ruins the model...
2026-02-23 17:59:29
0
Biopedialogy :
Good video :) I was wondering why we don't use XGboost over random forest?
2026-07-30 09:03:47
1
Sam :
why not just model through OLS or MLE?
2025-12-07 00:27:05
0
Vale :
That’s well explained and intuitive, however I think it misses the crucial information: how you measure the entropy for the dataset in this example, in other words how it know that the leaf have been reached?
2025-11-28 19:54:50
1
g746787454875 :
Is the same as CART
2026-01-12 02:25:19
0
JC Smitherson :
Github link for the code? 👩💻
2026-01-05 21:07:25
0
Secret Rap Battle :
Let’s
2025-11-28 00:29:36
1
Ani or Amination :
i am a huge fan of you
2025-11-27 07:37:16
2
jmw337 :
👍👍👍
2025-11-27 08:57:16
1
Tatiana :
👍👍👍
2025-11-27 03:38:12
1
Denise :
😭
2025-12-04 11:51:18
0
aladdin :
♥♥♥
2025-11-27 21:30:06
0
CtrlAltFun :
kinda ID3 alghoritm
2025-11-28 06:31:01
1
To see more videos from user @lerabyte, please go to the Tikwm
homepage.