Neocortex 🧠

Search

Search IconIcon to open search

Classification

Last updated Sep 21, 2022 Edit Source

Classification is the problem of guessing to which class it belongs(often represented by the symbol $\omega$) given a data point. For this problem, it is quite common to calculate the posterior probability of a data point for each class and classifying that point as the group with the largest probability.

Given this definition, a classifier can be defined in the following ways:

However, since the posterior is hard to estimate in most cases, it is quite common to use the Bayes’ theorem. It maps to classification like so:

$$ p(\omega | x) = \frac{p(x|\omega)p(\omega)}{p(x)} $$

In this formula the terms are:

How is class distribution calculated

It is guessed using contextual information on the data. More often than not, it is a gaussian distribution.


Interactive Graph