Evaluate a hybrid deep learning model

Hi, I'm currently doing a project using CNN as features extractor and train an SVM to classify the features extracted. The confusion I'm currently having now is I do not know how to determine whether the model is undertraining, overtraining or having generalisation problem. The example i was referring uses confusion matrix to evaluate the classifier but is that sufficient to tell the user that the model is accurate and it does not have the problems mentioned? Is there any other ways to validate the model?

Answer 1

[Sai Bhargav Avula]

Hi,

A confusion matrix is a table that is often used to describe the performance of a classification model (or “classifier”) on a set of test data for which the true values are known. Using this you can evaluate the accuracy, precision and recall which are the metrics for any classification model.

Hope this helps!