This is amazing. Using these loss functions for last 2 years but neither anyone taught me this nor I found such derivation on the internet. Love you for this video.
Hey man. Thanks a lot for considering my request. Really appreciate. Can u pls explain how the sigmoid curve is derived in logistic regression as we can probably have many of them. Like we use ols to find the best fit line in linear regression. Similarly how the sigmoid curve is found
It helped me understand the concept thanks. I find the way you explain is little fast, can you please slow a down a little bit such that we can digest.
In this video, I've explained linear regression in great detail. Let me know if you still have doubts after watching this one. czcams.com/video/fnDO1s4fzi4/video.html
Hey! Good attempt. You still have a bit of things to correct for. Use presentation slides for clarity; or use writings that are legible and consistent for the same clarity issue. Give some more background information for the theories: examples, more theoretical story .. etc
Hey man. Thanks a lot for considering my request. Really appreciate. Can u pls explain how the sigmoid curve is derived in logistic regression as we can probably have many of them. Like we use ols to find the best fit line in linear regression. Similarly how the sigmoid curve is found
Glad this video helped you. Now here's the thing... I think you have confused the shape of the sigmoid curve with the shape of the boundary in case of logistic regression. As long as we are passing a linear function (h(x)) inside the sigmoid function we can only get linear boundaries (a line in 2D, a plane in 3D and so on). So in logistic regression also we are finding the best fit line. The difference is that...In linear regression, the best fit line is that line from which the squared distance of all points is minimum but in logistic regression, the best fit line is that line which acts as the best boundary between the 2 classes. The sigmoid function (non-linear) just helps with the probability and does not affect the linearity of the boundary! In fact, this the reason why logistic regression can not solve the xor problem.
@@NormalizedNerd so do u mean we find the best fit line in logistic regression and then apply sigmoid function in order to get the linear boundary between the two classes!
@@NormalizedNerd am still confused. In almost every video I have seen they have explained about restricting the values between 0 and 1 by applying sigmoid.
Bro..This is THE EXACT video i was looking for. Literally explained like i am a 5 year old.
Thanks bro :D
This is amazing. Using these loss functions for last 2 years but neither anyone taught me this nor I found such derivation on the internet. Love you for this video.
Glad to be able to make these videos...Keep supporting ❤️
Hey man. Thanks a lot for considering my request. Really appreciate. Can u pls explain how the sigmoid curve is derived in logistic regression as we can probably have many of them. Like we use ols to find the best fit line in linear regression. Similarly how the sigmoid curve is found
It helped me understand the concept thanks. I find the way you explain is little fast, can you please slow a down a little bit such that we can digest.
Thanks for your feedback. I'll try to follow that. Keep supporting.
fan of your amazing explanation dear
10:08 excuse me sir why is it not mx + c + errorterm?
You are great man !!!!!!!! If possible please make a video on Expectation-Maximization and how k-means clustering is related to it.
Suggestion noted! Keep supporting :D
Great explanation
Also can u pls make a quick video on linear regression assumptions as many youtube video confuses a lot
In this video, I've explained linear regression in great detail. Let me know if you still have doubts after watching this one. czcams.com/video/fnDO1s4fzi4/video.html
Very helpful!
This is awsome ... great job :-)
From 10:49+, I think you mean to say Joint Density Function and not Probability.
Good video!!
Hey! Good attempt. You still have a bit of things to correct for. Use presentation slides for clarity; or use writings that are legible and consistent for the same clarity issue. Give some more background information for the theories: examples, more theoretical story .. etc
Feedback noted!
You should be citing some resources my dude. Your accent is not enough :P
Hey man. Thanks a lot for considering my request. Really appreciate. Can u pls explain how the sigmoid curve is derived in logistic regression as we can probably have many of them. Like we use ols to find the best fit line in linear regression. Similarly how the sigmoid curve is found
Glad this video helped you.
Now here's the thing...
I think you have confused the shape of the sigmoid curve with the shape of the boundary in case of logistic regression. As long as we are passing a linear function (h(x)) inside the sigmoid function we can only get linear boundaries (a line in 2D, a plane in 3D and so on). So in logistic regression also we are finding the best fit line. The difference is that...In linear regression, the best fit line is that line from which the squared distance of all points is minimum but in logistic regression, the best fit line is that line which acts as the best boundary between the 2 classes. The sigmoid function (non-linear) just helps with the probability and does not affect the linearity of the boundary! In fact, this the reason why logistic regression can not solve the xor problem.
@@NormalizedNerd so do u mean we find the best fit line in logistic regression and then apply sigmoid function in order to get the linear boundary between the two classes!
@@SumitKumar-uq3dg No, I wanted to say that the best fit line (in the context of logistic regression) is the boundary line.
@@NormalizedNerd am still confused. In almost every video I have seen they have explained about restricting the values between 0 and 1 by applying sigmoid.
@@SumitKumar-uq3dgYes that is true. Restricting the values between (0,1) gives the sense of probability. But where is your exact confusion?