Decision Tree (Basic Intuition - Entropy, Gini Impurity & Information Gain) | NerdML

This video will help you to understand about basic intuition of Entropy, Information Gain & Gini Impurity used for building Decision Tree algorithm. We will mathematically solve the problem. I have divided Decision Tree tutorial into several parts which will cover basic intuition, Classification problem solving & regression problem solving.
Below topics are explained in this video:
1). Agenda (00:40)
2). Introduction to Decision Tree (01:27)
3). Entropy (03:40)
4). Information Gain (07:53)
5). Gini Impurity (11:12)
Do subscribe to my channel and hit the bell icon to never miss an update in the future:
/ @nerdml
Please find the previous Video link -
Support Vector Machine Kernel Trick (Part - 4) | NerdML : • Support Vector Machine...
For connectivity with Machine Learning enthusiasts & professionals hit the link :
/ ntirawen1
Prerequisites
Basic understanding of Linear Algebra, Probability, Calculus, Matrix & Python programming including pandas, numpy, scikit learn & some visualization tools.
------------------------------------------------------
Creator : Rahul Saini
Please write back to me at [e-mail] for more information
Instagram: / 96_saini
Facebook: / rahulsainipusa
LinkedIn: / rahul-s-22ba1993
#DecisionTree, #MachineLearning, #NerdML, #Entropy, #InformationGain, #GiniImpurity, #Mathematics

[@ronitroy2887]

Hey buddy, I was waiting for this video. This is really gud. Keep up the good work👍

[@NerdML]

Thanks buddy!!

[@thangtran145]

Great video, subscribed! Clear, concise, friendly, and easy-going !

[@NerdML]

Thanks for liking it Happy Learning!!

[@raghunathanp3734]

Brilliant stuff .. keep going .. very helpful for my ML exam in 2days

[@raghunathanp3734]

One more doubt the high value of each of the entropy, Gini, information gain indicates what?

[@NerdML]

All the best

[@NerdML]

Information gain & entropy value is inversely proportional to each other so, at the same time both of these values can not be high. (Information gain should be high & Entropy should be low for ideal condition.)

[@lavanyarao5650]

Thank you so much

[@NerdML]

My pleasure

[@arpitcruz]

Buddy want tell you that there are hell lot of channel for data science .. but no one is better than u the way u explain

[@NerdML]

Thanks, keep sharing the content with your community

[@siddheshpowar4208]

Nice explanation ✌️...all the things are nicely presents 👌.

[@NerdML]

Thanks dude👍🏻 Keep supporting

[@naivelearner6357]

amazingly explained...

[@NerdML]

Oh thanks😊

[@mohinikumari819]

👍

[@in5minutes27]

Thank u so much

[@NerdML]

Keep supporting 👍

[@akshara_K.]

Awesome video with detailed explanations. Can you suggest other topics/concepts that need to be learnt for Data science? If they are already included in your channel, that would be great.

[@NerdML]

Thanks Akshara, glad you liked the content Recently I have uploaded a video on “Complete Roadmap for a learning Data Science from scratch” You can watch that

[@akshara_K.]

@@NerdML Thank you very much! Much appreciated!

[@NerdML]

Pleasure is mine!!

[@hannahnguyen4865]

Thank u

[@NerdML]

My pleasure

[@ayushsingh-qn8sb]

awesome explaination

[@NerdML]

Thanks dude, keep supporting 👍

[@debabratasahoo2588]

Hlw sir , your content is osm. I much clarify throughout your good teaching way. thanks sir

[@NerdML]

Thanks to you for liking it Happy Learning!!

[@VarunSharma-ym2ns]

Grt session 👌

[@NerdML]

Thanks Varun!!

[@topperbizzare564]

*NerdML* 💕💐 execellent video 💕 i saw this is too good 💕 thanks for this 💕

[@NerdML]

Thanks mate!!

[@tymothylim6550]

Thank you very much for the video! Really clear and helpful:)

[@NerdML]

Thanks and keep supporting

[@patrickmckenna6520]

Where do nodes E, F and G come from? Is each node not based off an attribute?

[@NerdML]

Sorry for late response Node E, F & G are branches of Impure Sub-tree B & C. As Node B is having 2 Yes & 4 No so further we will divide this node in D & E and assign 2 Yes, 0 No to D and 0 Yes, 4 No to E so that we can reach at pure leaf node. Definitely, each & every node is based off an attribute that's why we are finding entropy of each node & taking valid node for further processing. Hope this make sense to you!

[@uditarpit]

Entropy measures impurity!!

[@vaddadisairahul2956]

May I know why is the value of entropy shown in bits? anything we should know about it?

[@NerdML]

See it's a standard practice to keep it in bits (binary digits) as due to choice of the base-2 logarithm

[@NerdML]

If we use normal logarithm instead of base-2 then unit will be nats

[@BalaMurugan-cm6ev]

Wow . . Nice Explanation. Even if we calculate Gini Impurity, we will calculate Information Gain. . . Am I right?

[@NerdML]

Yes you can there is minor change in both the entropy and gini index

[@kaustubhdwivedi1729]

Bas ek baat bolna chahta hoon ki MAJA AAYA !

[@NerdML]

Shukriya🙏🏻

[@vaishnavikalidass1544]

Finally I understood these 😂😂😂

[@NerdML]

Good to know...keep supporting 👍

[@vaishnavikalidass1544]

@@NerdML can u post more videos like step by step preparation to Data science

[@NerdML]

Sure I will whenever I get some free time.