Given a data set of 50, 000 images (32x32 pixels) that belong to 10 categories, say, the 10 digits, how many parameters in your classifier for hand-written digits that you can comfortably entertain if you want to train (a) a discriminative classifier or (b) a generative classifier?
I need help solve this question with argument.
5 freelanceria on tarjonnut keskimäärin %project_bid_stats_avg_sub_26% %project_currencyDetails_sign_sub_27% tähän työhön
Hi, there is no unique answer. I can help you understand the problem with an in-depth explanation. I am a phd computer engineer with great experience in deep learning and I work in the context of computer vision.
I have been doing Neural Networks and Machine Learning since 3 years now. My Final Year Thesis is also on Machine Learning. But your question is a bit unclear. I would require more details before answering.