Yes, there are all kinds of tasks where the appropriate solution is to use a DNN for much of the learning (either directly learning the correlations or as trans

Yes, there are all kinds of tasks where the appropriate solution is to use a DNN for much of the learning (either directly learning the correlations or as transfer learning from some large-data self-supervised task) and then, once you have the results of that DNN inference, work with these methods - apply PCA for interpreting the resulting vector, or to separate out specific dimensions to expose them for adjustment in some generative task; or perhaps the best way for the final decision is a kNN on top of the DNN output, etc.