139 - The topology of deep neural networks, designing your model.

[@ahmedbalaazi7806]

Thank you so much for the huge effort you are doing..

[@nagasudha6928]

Thanks a lot Sreeni in making us understand the AIML with which we got huge confidence on starting of the research

[@DigitalSreeni]

Glad to hear that :)

[@samarafroz9852]

Big fan of you

[@samarafroz9852]

Awesome work sir 👍

[@DigitalSreeni]

Thank you! Cheers!

[@veenags2735]

Awesome work.can u pls demnstrate stacked generalusation ensemble model of cnn

[@Mojtaba-Sirati-Amsheh]

Thanks

[@haowenwei3256]

Hi sir, I have a question. Why the input data dimension is 32 * 32 * 3.. Where is that 3 comes from. Thank you very much :)

[@medblgsec]

it's for the three channels RGB, because the dataset presents some colors and not only greyscale images !

[@sumitlahiri209]

3 comes from the fact that there are 3 channels (types/components) of data for each gridcell in 32x32 , i.e the RGB (colorful images that it) channels. On the contrary, MNIST Digit dataset has only 1 channel so, the dimensions are 28x28, 1 (they are black and white)

[@Jango1989]

A great video but I feel like you didn't actually explain how and why we design the topology of a model; you only showed how the model's performance improved as the topology grew deeper.