1

This CMU Machine Learning Text Book says

when we are interested in learning some target function $f : X → Y$, we can more generally learn the probabilistic function $P(Y|X)$.

there are 2 forms above to represent a Machine Learning model.

is reasonable to say kNN is an example of this form

$f : X → Y$

while SVM is an example of another form

$P(Y|X)$

could someone please double check my understanding about these 2 forms?

JJJohn
  • 683
  • 2
  • 11

1 Answers1

0

$P(Y|X)$ is a conditional probability representation where Y can be any output given input X. kNN can also be represented in this form please check this answer!. SVM can also be represented similarly.

$f:X\to Y$ denotes a function mapping between input X and output Y. which means there is a function that maps X to Y and our aim in the machine learning or statistical problem is to learn this function.

Tim
  • 108,699
  • 20
  • 212
  • 390
Abinav R
  • 1
  • 2