Видео

nice video btw

Thank you

Thank you very much!

Thank you very much! Very helpful examples too. I have a question though, is the bias always 1?

Thanks

Good explanation!

This is well-explained, thank you sir!

Very well explained, thanks.

Thank you so much. Your explanation was clear and helped me a lot.

I think that it is very good explanation I have ever seen

Thank you

did u fart 13:14

Very good, this is by far the best explanation of this matter that i have come across. I want to use statsmodels.tsa.seasonal in python and i was having a hard time understanding what a seasonal value of x really means. With this content i understand it much better!

You are doing amazing explanation with the clear understanding and knowledge you have. Thanks

On the minimal twice a year delivering the workshop, who arranged the workshop? Us/University or Nvidia? And could tell how you arranged to deliver them, thanks

Worst explanation of my life

Great Video, Thank You!

Thank you for the video :) helped me wrap my head around it

)0

You da best 🎉

Thank you very much 🎉

Is it just me that think those API really full of bad smells? For starters, why would someone write two ways to do the same thing? Thats making stuffs more complicated for no real gain in the send. That way of chaining functions in functional programming really hurt my ways when compared to classical chaining from functional programming.

Great explanation!

In the else if statement, there is a negative sign missing, it should be e ^ (-dE / T). This term represents the probability of accepting the new state, so it makes sense that e ^ (-dE / T) approaches 0 as dE increases and T decreases.

Great explanation, very well put thank you

I like your teaching style sir,it is 100 times good than my class teacher

Amazing explanation! Saw Andrew Ng's course but was having trouble understanding why exactly do residual connections work, now it is crystal clear. Thanks a lot!

Respected sir, Your course is the best Ai course I have stumbled upon and I am truly grateful to learn from you. But I would like to know where I could get acesss to the rest of the videos as there are many videos and chapter missing from this playlist. Thank you :)

I've been through 5 videos so far, and it finally clicked in my mind thanks to this proof. Thank you so much.

😢شرحك يبدو جيد ولكني لا أفهم الا العربية 😔

just finished this playlist and I can say its a gem on youtube

useless video

you did not explain the basic architectural difference....

As many comments note--this is a great quick explanation of the use of 1 x 1 convolutions in multiple contexts. I had watched 2 other 15+ minute videos and understood the math behind it. But the application and how it can be useful was missing in those concept videos. THANK YOU, Badri!

This is amazing. Thank you so much.

data or github link sir

this is cool, no one covered this from such a perspective , thank you! I wish you would show the things that work as kernels in our visual process. I wonder: do we have copies of many kernels for convolution? Because obviously it's unlikely the kernel "moves", like when we perform convolution on a computer. So there must be hundreds of little neural kernels that are exact copies of each other!!!

Exercise: Apply Uninformed Search Strategies to identify optimal path

Thank you so much, very well explained.

def fully_connected(input): x=tf.keras.layers.Dropout(0.2)(input) x= tf.keras.layers.Dense(512,activation='relu',kernel_regularizer='l1_l2',activity_regularizer='l1')(x) x=tf.keras.layers.Dense(256)(x) x= tf.keras.layers.BatchNormalization()(x) x=tf.keras.layers.Activation("relu")(x) output=tf.keras.layers.Dense(512,activation='softmax')(x) return tf.keras.Model(input, output)

this is soooooooo gooood. Highly recommend to anyone who wants to understand the # of parameters in CNN

Learnt something. Thank you professor

I have a question, in the example at last where C has leaf nodes x and y. Can we prune the tree with roots x and y?

Finally someone who can explain it simply. Thank you very much.

For Convolution layer : Number of Params = (filter width × filter height × input channels + 1) × number of filters For Fully Connected layer: Number of Params= (current layer neurons c * previous layer neurons p) + c.

JAY NEPAL

life saver ! may Allah give u health and strength sir great explanation !!!!

Good Job Professor

This can be used in trading

great explanation