Generating a supervised dataset
irfanfadhullah opened this issue · comments
Hello, thank you for the information on how to generate the unsupervised dataset. I want to ask another question, so I decided to open a new issue.
Is the process the same for generating a supervised learning dataset? For example, your previous paper "RSSI-Based HBF Design with Deep Learning" used a supervised dataset.
If the process is not the same, could you provide the information on how to generate that dataset?
Thank you
Yes, the process is same.However, since it is supervised learning, there would be the optimal values as target. Thus, the dataset consist of RSSI, CSI and optimal values.
So, in order to get the optimal values of HB that using perfect CSI, we need to find it by using the steps in the Section II.B in the "RSSI-Based HBF Design with Deep Learning" paper right?
Yes or you can use III.C in "Unsupervised Deep Learning for Massive MIMO Hybrid Beamforming".