Sigmoid for multi-label not multi-class

How sigmoid will be used here, suppose i have a3_1,a3_2,a3_3 then what is criteria for choosing that bus is there or car is there.

You will be using softmax for more than binary output and I guess you will be learning about it soon enough :slight_smile:

Hello @Aman6,

As the name of the video suggested, it is about a classification with multiple outputs, so you can have more than one outputs. In that slide, each sigmoid is responsible for one output. For how the sigmoid is used, I will refer you back to the lectures that discussed how one sigmoid is used in a logistic regression problem. It is the same use.

Good luck.