I have selected two deep learning models CNN and sae for data analysis of a 1 d digitized data set. I need to justify choice of these two dl models in comparison to other dl and standard ml models. I am using ga to optimize hyper parameters values of the two dl models. Can you give some inputs for this query.thanks.
Sergei V. Dolin
Hello, Rajan Vohra.
Typically, the rationale for choosing a model can be training time, prediction time, and the value of the metric itself, either on a validation set or cross-validation, depending on what you are using. It is better, of course, to use more than one metric for indicators, as well as an error matrix along with completeness and accuracy, or simply F1 or F1-beta, depending on the problem you are solving.
0 votes
0 thanks
- Badges
- Science topic
- Similar topics
- Analytical Chemistry
- Column
More Rajan Vohra's questions See All
Similar questions and discussions