According to my knowledge dropouts are used to drop some neurons but here number of neurons are same as before using dropouts. Why so?
And how it Dense layer working like flatter layer have 8 neurons and how Dense layer will convert it into 8 neurons?
Yeah you got it right in some way. Actually , the thing dropout does is that it doesn’t actually drop some neurons , its just lower there weights value , such that they have somewhat a no role in working. You can think it as changing there weights to 0. So , in this way you will be having the same shape , but the values and computations in the network will reduced and enhanced.
Dense Layer is the most basic and simple layer in deep learning. Its just forms a simple ANN structure , which will have as input 8 nodes and output as 10 nodes. Now in this ANN ,All the calculations will be made using weights and biases , same as a feed forward network does.
I hope this helped you in understanding the thing.
Thank You and Happy Learning 
.
Thanks a lot. It is clear to me like never before.
Its really good to hear that.
I would request you to kindly mark this doubt as resolved in your course doubt’s section and also do provide your valuable feedback as it helps us to improve this platform.
Thank You and Happy Learning .
