0

Consider the probkem of binary classification using Naive Bayes Classifier. You are given two dimensional features and categorical conditiobal distributions in the tables below.The question and related table is given in the picture. I applied bayes rule as: $\cfrac{P(X_1=-1,X_2=1|C_1)}{P(X_1=-1,X_2=1)} = \cfrac{P(X_1=-1|C_1)P(X_2=1|C1)P(C_1)}{P(X_1=-1|C_1)P(X_2=1|C1)P(C_1) + P(X_1=-1|C2)P(X_2=1|C2)P(C_2)}$

From the table, I think:

$P(X_1=-1|C_1)=0.2$

$P(X_2=1|C1)=0.1$

$P(X_1=-1|C2)=0.3$

$P(X_2=1|C2)=0.6$

I couldn't make anything about $P(C_1)$ and $P(C_2)$. May you please help me? Thanks in advance for your replies.

Edit:Picture is added!

kursat
  • 13
  • 3
  • You didn't attach any picture. – Tim Jan 31 '22 at 15:07
  • @Tim, picture has been added. Sorry for that mistake. – kursat Feb 01 '22 at 07:03
  • @kursat Forgive me if I misunderstand, but the "...two classes are equally likely..." should imply that $P(C_1)=P(C_2)=0.5$. Also, I think there is a typo on the LHS part of the equation where you apply Bayes' rule (missing $P(C_1)$) – stochastic13 Feb 03 '22 at 13:03

0 Answers0