Share your linear regression and logistic regression notebooks & blog posts here.
Have you ever heard about backpropagation? if so, take a look at this notebook and find out how PyTorch is implementing it. If not and you want to learn then this could be the perfect notebook for you
your feedback and supports are highly appreciated, and if you want to contribute please feel free to.
If you find it useful don’t forget to like my notebook, Thanks in advance. Learning with sharing.
Thanks @aakashns and Jovian team for your efforts!
Hi folks, check out my 2nd assignment on insurance prediction using regression technique via PyTorch, I am open to your suggestions and corrections.
click here to view notebook.
Your notebook is incorrect: your predicted value is among the inputs feed into the model.
so, what i need to do to correct that now
Well, remove this column (
charges) from your inputs.
Yeah! Now i got you…i am giving output column in inputs… Right???
You basically predict the answer from the question like this
Something like “What is the name of George?”
yeah i understood sorry for the Mistake…iam just a beginner Not a Pro
This is my 2nd assignment… make sure to like my notebook.
After, this assignment I learn many new things which might be helpful our daily life
Hello guys! Please check out my work and feel free to criticize. Link: https://jovian.ai/light-ti-man/02-insurance-linear-regression-with-jovian
Hi friends, I am santosh. I was doing assingnment on images and logistic regression. In the middle I got this error.
can anyone help me with this.
Hey @santosh-ui the colab notebook is private can you commit it to jovian and share that link?
Hi guys, Sharing the link to my notebook.
Thank you guys for sharing: you helped me to understand that I had to replace the cross entropy loss as explained here
Finally, my assignment 2 notebook
The statement about being unable to use cross entropy doesn’t make much sense to me. You can’t apply classification-related function to regression.
This would be different if you wanted to for example make some ranges into which single people’s costs would fall into. You could then split the dataset into distinct classes. Since it would be single-label classification (one person can’t have 2 different insurance costs) you could use cross entropy without a problem then.
So my statement makes sense and we’re just saying the same thing with different words!))
It doesn’t make sense in the meaning that it’s stating something obviously incorrect/unapplicable to this task.
It’s as if you said you learnt that you can’t use spectrogram here (because there’s no sound or any signal), you can’t use PHP (because you use python) or you can’t use a stick to train the model.
I can’t imagine how many things we could learn this way, if we applied the unapplicable