Stack Exchange Network
Stack Exchange network consists of 181 Q&A communities including
Stack Overflow, the largest, most trusted online community for developers to learn, share their knowledge, and build their careers.
Visit Stack Exchange
Artificial Intelligence Stack Exchange is a question and answer site for people interested in conceptual questions about life and challenges in a world where "cognitive" functions can be mimicked in purely digital environment. It only takes a minute to sign up.
Sign up to join this community
Anybody can ask a question
The best answers are voted up and rise to the top
7 months ago
I have multiple DNN that can extract features vector from images.
Those can be used for two main goals:
Use them for transfer learning ang faster trainings
Use them as feature extractors, and train only additional, small "heads" (in particular if my dataset is really small)
But how can I choose which trained network to use?
For the moment, I use:
A UMAP projection on a labelled dataset (that the networks have never seen), and a compare visually how well my various classes are "clustered".
A "Raw 10-NN Mean" score. For each point, I count how many of their 10 nearest neighbours are of the same class. I Divide by 10 to get the score.
A "Class 10-NN Mean" score. Same as before, but each class has the same weight, regardless of their proportion in the dataset.
Are there any other interesting metrics I could use to compare my feature extractors?
Aug 5, 2022 at 13:10
log in to answer this question.
By clicking “Accept all cookies”, you agree Stack Exchange can store cookies on your device and disclose information in accordance with our
Accept all cookies
Necessary cookies only