1
$\begingroup$

Just curious to know the meaning contained in the latent variables used in the machine learning applications ?

Basically, I am confused if latent variables present the same essence as the model parameters ? In Parameters vs latent variables, @alberto suggested that parameters are fixed while latent variables are essentially random variables that are not observed directly.

My point is that even a parameter is a random variable (its just that we do not know the value of this parameter). Surely, we know the distribution that it is coming from and using this distribution as a prior information, we try to estimate the value that fits the training data well. Am I going well with this concept ?

$\endgroup$
  • 1
    $\begingroup$ Parameters are constant across all data points while latent variables can vary. Yes, parameters can be treated as random variables too, which was pointed out in the other thread. $\endgroup$ – shimao Mar 16 '18 at 4:16
  • $\begingroup$ @shimao I've copied your comment as an answer. If you would prefer to write your own answer, please let me know and I'll delete mine. $\endgroup$ – Reinstate Monica Jul 28 '18 at 16:40
1
$\begingroup$

Parameters are constant across all data points while latent variables can vary. Yes, parameters can be treated as random variables too, which was pointed out in the other thread.


I've copied @shimao's comment as an answer because the comment is, more or less, an answer to this question. We have a dramatic gap between answers and questions. At least part of the problem is that some questions are answered in comments: if comments which answered the question were answers instead, we would have fewer unanswered questions.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.