Deep Learning with PyTorch Live Course - Image Classification with CNNs (Part 4 of 6)
video description
Date: 2022-03-14
Related videos
Comments and reviews: 9
Sam
At 1:31:37 and also earlier, Aakash says that we can pick the max. probability. Actually, the output from the model is not really probability. It is a floating point number (negative and positive) and it is rather the softmax layer (embedded into the cross-entropy loss function) that turns these positive and negative numbers into a probability (+ve) value that sums to 1. But picking the max. of such a output is still fine, since that is just the number that gives the max. probability when passed through the cross-entropy.
reply
At 1:31:37 and also earlier, Aakash says that we can pick the max. probability. Actually, the output from the model is not really probability. It is a floating point number (negative and positive) and it is rather the softmax layer (embedded into the cross-entropy loss function) that turns these positive and negative numbers into a probability (+ve) value that sums to 1. But picking the max. of such a output is still fine, since that is just the number that gives the max. probability when passed through the cross-entropy.
reply
Odejide
Good day sir, I got to know about this program during the third week, i tried to register via the link provided but i later realized the registration has been closed. I just want to know if people in my category will have opportunity to participate in Kaggle competition scheduled at the end of this program.
I have watched all the previous videos and did all the previous given assignment but i did not submit since am not eligible.
Let me know your thought sir.
Thanks in advance.
reply
Good day sir, I got to know about this program during the third week, i tried to register via the link provided but i later realized the registration has been closed. I just want to know if people in my category will have opportunity to participate in Kaggle competition scheduled at the end of this program.
I have watched all the previous videos and did all the previous given assignment but i did not submit since am not eligible.
Let me know your thought sir.
Thanks in advance.
reply
Nikolay
Is there such thing as junior machine learning engineer? Is it possible to get your very first job as programmer straight into ML? And that on the base of being self taught, studying for about 6 months, about 2 hours per day. Also moving from totally different field and starting your studies from the basics of Python.
reply
Is there such thing as junior machine learning engineer? Is it possible to get your very first job as programmer straight into ML? And that on the base of being self taught, studying for about 6 months, about 2 hours per day. Also moving from totally different field and starting your studies from the basics of Python.
reply
Multitudes
This has been one of the best videos so far. Aakash did a great job in explaining the convolution and overfitting problems. Really enjoying the series so far! I like how we came here gradually applying different techniques hands on and understanding the limits of each one!
reply
This has been one of the best videos so far. Aakash did a great job in explaining the convolution and overfitting problems. Really enjoying the series so far! I like how we came here gradually applying different techniques hands on and understanding the limits of each one!
reply
Shikhar
your dataloader normalize the image in the range(0,1) so when you plot them why are they giving the real image . I mean to say why the image didn't get distorted after changing the pixel values.
reply
your dataloader normalize the image in the range(0,1) so when you plot them why are they giving the real image . I mean to say why the image didn't get distorted after changing the pixel values.
reply
saurrav
How to make two models and predict the outcome.....like in training dataset contain two type data normal and differents......so while predicting how to known it's normal or different
reply
How to make two models and predict the outcome.....like in training dataset contain two type data normal and differents......so while predicting how to known it's normal or different
reply
jdragon
what good is a free course if they dont teach basic math operation , nothing major was covered , but its better than nothing so use it to fullest and learn urself
reply
what good is a free course if they dont teach basic math operation , nothing major was covered , but its better than nothing so use it to fullest and learn urself
reply
ba
hey, could you do more of these in class kaggle pytorch competitions. I just got to know about it so late and there is just one day to go to compete :(
reply
hey, could you do more of these in class kaggle pytorch competitions. I just got to know about it so late and there is just one day to go to compete :(
reply
Amit
One thing, I want to clarify 1:47:33 that these are not proteins but cell organelles and some are composed of proteins and other macromolecules
reply
One thing, I want to clarify 1:47:33 that these are not proteins but cell organelles and some are composed of proteins and other macromolecules
reply
Add a review, comment
Other channel videos
