Entropy dalam machine learning dan AI

 Entropy

Assume we have P predictors and K classes. Suppose we select the kth predictor and split a region along the threshold .

We can assess the quality of this split by measuring the entropy of the class distribution in each newly created region by calculating:

Note: We are actually computing the conditional entropy of the distribution of training points amongst the K classes given that the point is in region r.

The same plot as last time, with R1 containing only blue dots and R2 containing both blue dots and orange triangles.


The entropy calculation here yields a value of 1.38, compared to a misclassification rate of 0.38 and a Gini index of 0.47.

We can now try to find the predictor p and the threshold tp minimizes the weighted average entropy over the two regions:

Where Nr is the number of training points inside of region Rr.


Comments

Popular posts from this blog

create image slider using phyton in web

Tahukah kamu Algoritma Genetika dan Penerapannya dalam Industri

create animated futuristic profile card using html+css+js

CRUD SPRING REACTIVE WEBFLUX +Mongo DB

Top 7 Digital Transformation Companies

100 perusahaan perangkat lunak (software) populer dari Eropa dan Amerika yang memiliki kehadiran atau operasional di Indonesia.

TOP 8 Framework Populer menggunakan bahasa .NET

Python Date and Time Manipulation

TOP 5 Trends Programming 2024

Daftar Kata Kunci (Keyword) dalam Bahasa Pemrograman Python