Python - Splitting Criteria Entrophy | Python Courses in Tamil | Skillfloor - video Dailymotion

Skillfloor

Understand the concept of Entropy, a fundamental splitting criterion used in Decision Tree algorithms, with this easy-to-follow tutorial in Tamil!   Our Website: Visit 🔗 http://www.skillfloor.com  Our Blogs: Visit 🔗 https://skillfloor.com/blog/  DEVELOPMENT TRAINING IN CHENNAI https://skillfloor.com/development-training-in-chennai  DEVELOPMENT TRAINING IN COIMBATORE https://skillfloor.com/development-training-in-coimbatore  Our Development Courses: Certified Python Developer Visit 🔗https://skillfloor.com/certified-python-developer Certified Data BASE Developer Visit 🔗https://skillfloor.com/certified-data-base-developer Certified Android App Developer Visit 🔗https://skillfloor.com/certified-android-app-developer Certified IOS App Developer Visit 🔗https://skillfloor.com/certified-ios-app-developer Certified Flutter Developer Visit 🔗https://skillfloor.com/certified-flutter-developer Certified Full Stack Developer Visit 🔗https://skillfloor.com/certified-full-stack-developer Certified Front End Developer Visit 🔗https://skillfloor.com/certified-front-end-developer  Our Classroom Locations: Bangalore - https://maps.app.goo.gl/ZKTSJNCKTihQqfgx6 Chennai - https://maps.app.goo.gl/36gvPAnwqVWWoWD47 Coimbatore - https://maps.app.goo.gl/BvEpAWtdbDUuTf1G6 Hyderabad - https://maps.app.goo.gl/NyPwrN35b3EoUDHCA Ahmedabad - https://maps.app.goo.gl/uSizg8qngBMyLhC76 Pune - https://maps.app.goo.gl/JbGVtDgNQA7hpJYj9  Our Additional Course: Analytics Course https://skillfloor.com/analytics-courses https://skillfloor.com/analytics-training-in-bangalore Artificial Intelligence Course https://skillfloor.com/artificial-intelligence-courses https://skillfloor.com/artificial-intelligence-training-in-bangalore Data Science Course https://skillfloor.com/data-science-courses https://skillfloor.com/data-science-course-in-bangalore Digital Marketing https://skillfloor.com/digital-marketing-courses https://skillfloor.com/digital-marketing-courses-in-bangalore Ethical Hacking https://skillfloor.com/ethical-hacking-courses https://skillfloor.com/cyber-security-training-in-bangalore   #entropy #decisiontree #pythontutorial #machinelearning #tamilcoding #skillfloor #datascience #pythonintamil #mlalgorithm #splittingcriteria #classification #datasciencetamil #machinelearningtamil #pythonprogramming #mlmodels #techeducation #codingintamil #featureselection #pythoncourse

Transcript

00:00Hello everyone, in this video, we are splitting criteria, entropy.

00:08So, entropy based upon each tree, we will split up.

00:12So, entropy is the inner criteria.

00:17Randomness and oddness in our data, we will calculate.

00:21Normally, we compare the entropy, we will work slower.

00:26We will work closer to the entropy.

00:28Then, we will calculate the entropy and probability based upon each tree.

00:32Guinea split index, impurity and randomness.

00:34But entropy is the probability and logarithmic.

00:37So, if we combine it, we will provide the decision.

00:41If entropy is equal to zero, we will calculate the pure split.

00:46We calculate the weighted average in our data.

00:50We calculate the information in our data.

00:53So, we calculate the guinea split index.

00:56We calculate the entropy of entropy.

00:58We check the information gain.

01:00So, the information gain, we check the split.

01:02We check the split.

01:04We check the split index.

01:06In the guinea split index, we calculate the minimum value.

01:10We choose the minimum value.

01:12Correct?

01:13And the information gain, we choose the highest value of the column.

01:18That's the root node.

01:20So, the information gain formula is the E of S.

01:24That's entropy of complete data set.

01:27Okay?

01:28And minus weighted average into entropy of each and every features.

01:32So, each and every features based upon the information gain.

01:35So, each and every features based upon the information gain.

01:36We decide.

01:37First, we consider what data set.

01:41Outlook, that's weather.

01:43Then, temperature, humidity, windy.

01:45If we compare it, we don't have a very high level.

01:47We say that we have a target column.

01:49So, in the target column, we have a yes or no.

01:53So, in the s or no, we have a 9 data.

01:56No, 5 data.

01:58So, total of 14 data.

02:00So, we can analyze the probability.

02:02Okay?

02:03So, if we analyze the probability,

02:05This formula is the entropy of the formula.

02:09Minus summation of probability of i

02:12into log to the base of probability of i.

02:15So, i enter the class.

02:17Okay?

02:18So, instance of play.

02:19Target column is classes.

02:20So, that is minus probability of s

02:23into log2 of probability of s.

02:26Then, minus probability of no into log2 of probability of no.

02:30Okay?

02:31We can calculate.

02:32So, probability of s is 9 by 14.

02:34Probability of node is 5 by 14.

02:36So, if we substitute it,

02:38we have 0.94.

02:39That is, we have 94% impurity in the data.

02:43Okay?

02:44So, first, we choose the root node.

02:47We calculate entropy.

02:49So, entropy we calculate.

02:51Like outlook.

02:52Like outlook.

02:53Like outlook.

02:54So, outlook.

02:55Basic.

02:56So, probability of s is 2 by 5.

02:57So, probability of s is 2 by 5.

02:59And rainy.

03:00So, sunny.

03:01So, sunny.

03:02There are total of 5 data.

03:03Rainy.

03:04There are 5 data.

03:05And overcast of 2 data.

03:06So, first, we will separate entropy.

03:07Sunny data.

03:08So, we calculate the entropy.

03:09Sunny data.

03:10So, sunny data.

03:11We pick up.

03:12So, here we know.

03:13We have no data.

03:141, 2, 3.

03:153 no data.

03:17And 2.

03:18Yes data.

03:19So, probability of s is 2 by 5.

03:225 is sunny data.

03:23Probability of s is 2 by 5.

03:25Probability of no is 3 by 5.

03:27So, what is the entropy?

03:29Minus probability of s.

03:31Into log 2 of s.

03:33Into plus probability of no.

03:35Into log 2 of 4.

03:37So, here the point is 971.

03:39That is, overcast column.

03:41So, overcast data.

03:42Overcast data.

03:44Overcast data.

03:45So, overcast data.

03:46There are 4 data.

03:47Actually.

03:48So, complete a yes.

03:49This is entropy equal to 0.

03:50Okay?

03:51Entropy equal to 0.

03:52Why?

03:53Completed we have yes data.

03:54Okay?

03:55Then, rainy data.

03:56Rainy data.

03:57That is what we calculate.

03:58So, this is the first weighted entropy.

04:00Okay?

04:01We calculate the entropy index.

04:03That is what we calculate.

04:05Probability of sunny into entropy of sunny.

04:08What we calculate.

04:09Probability of sunny into guinea value of sunny.

04:11And D.

04:12We calculate the value of 10.

04:13Then.

04:14Plus probability of overcast.

04:16Into entropy of overcast.

04:19Plus probability of raining.

04:22Into entropy of raining.

04:24So, this is completely calculated.

04:26We calculate the value of weighted entropy.

04:28Weighted entropy is 0.695.

04:30And then, we calculate the value of 8.

04:31That is the value of 8.

04:32That is the value of 9.

04:33So, the value of 0.9.

04:34So, we calculate the value of 0.9.

04:35data set value is 0.94 so 0.94 minus

04:40weighted entropy is 0.695 so the information gain of outlook is 0.2488

04:52so the temperature column is 0.2488 so the temperature column is 0.2488

05:03so hot cold mild to separate hot to get the entropy calculate mild and cool to calculate

05:09once after that we have weighted average in total 4 data so the hot data is considered

05:182s and 2no then probability of yes is 0.5 and probability of no is 0.5

05:27then mild data as to cool data as to calculate weighted entropy calculate

05:33so weighted entropy calculate probability of hot into entropy of hot plus probability of mild

05:39into entropy of mild plus probability of cool into entropy of cool

05:45so in the way calculate the weighted entropy calculate the bar

05:49if you calculate the pro information gain calculate

05:51so information gain of entropy of yes that is the entire data set minus weighted entropy

05:570.911 so 0.0328

06:01okay then similarly windy column go choose

06:05so windy column go same calculation so windy

06:08so windy level high normal

06:11end is split there so it is separated calculate

06:15high go normal go weighted entropy calculate

06:17information gain

06:18okay so this is third column

06:22fourth column

06:23fourth column

06:24fourth column

06:46the highest information gain is Outlook

06:49Outlook is the highest information gain

06:52Output is the root node

06:55Outlook is sunny, overcast and 20

06:58Overcast is the entropy is 0

07:01So it's splitting

07:03Sunny data and rainy data

07:06Then the second time decide

07:08Outlook

07:11Humidity column is yes or no

07:14Then windy data is strong or weak

07:18So we will create a root node

07:22Next further splitting

07:24Entropy calculate

07:26All the information gain higher

07:28Second node choose

07:30So this tree

07:32Entropy use

07:34Create

07:36In the next video

07:44Thank you

07:46To the next video

08:00Check it out

08:02All the information

08:04This tree

08:06To the next two

Python - Splitting Criteria Entrophy | Python Courses in Tamil | Skillfloor

Category

Transcript

Be the first to comment

Recommended