5.8 Backpropagation Algorithm Part 2 in Tamil

Рет қаралды 19,103

BECAUSE

Жыл бұрын

I have discussed about back propagation algorithm with example
Link for PPT
drive.google.com/drive/folder...

Пікірлер: 81

@indhumae5554 25 күн бұрын

I never seen like this video....crystal clear explanation... Great job...

@because2022 25 күн бұрын

Thank you so much. Keep learning and keep sharing.

@ramprakashjelari7385 5 ай бұрын

This is wonderful, and a very great serve to students community. How ever tough it is, if we learn by mother language, things get very easy to understand. This is absolute boon to students. Thanks a lot.

@because2022 5 ай бұрын

Thanks for your feedback. It means a lot to me. Keep sharing.

@ramprakashjelari7385 5 ай бұрын

Sure. Which e-mail id i can connect to you ?@@because2022

@CREATIVEINNOVATOR 7 ай бұрын

Clear explanation . Thank you very helpful

@because2022 7 ай бұрын

You are welcome!

@revathisathiyanarayanan1259 Жыл бұрын

Simple way of explanation

@because2022 Жыл бұрын

Thanks for the feedback.

@priyadharsini6654 Ай бұрын

Kadavuley ❤️❤️ nandri

@because2022 Ай бұрын

Most welcome. Keep dng well.

@monishav9634 28 күн бұрын

Great work sir🎉🎉🎉

@because2022 28 күн бұрын

Thanks a lot. Keep learning and keep sharing.

@dhineshkumarcse5064 Жыл бұрын

why are we not taking updated value at 20.15 ? we are taking 0.4 instead of 0.436

@because2022 Жыл бұрын

Good observation. We take upadates values in next forward propagation only.

@ratheeskan5959 8 ай бұрын

deivameee 🙏

@because2022 8 ай бұрын

Thanks.

@Just_Speaks 17 күн бұрын

12:12 what is the values for (out o -y ) sir

@because2022 17 күн бұрын

-0.4326. But if there are calculation mistakes, ignore it. Focus on concepts.

@harshavardhan.s6595 Ай бұрын

Super Sir 👍

@because2022 Ай бұрын

Thanks keep learning..

@user-pp5te1tx8v Жыл бұрын

sir in unit 5 is it enough study till this topic i mean till backpropagation

@because2022 Жыл бұрын

Sure a sola mudiyathu Kamesh.

@inevitable4104 8 ай бұрын

mam, backward propagation completely purijiruchu. but ithu enga gradient desent algorithm use aagathu.. Gradient desent oru video pannuinga mam

@because2022 8 ай бұрын

I have discussed about gradient descent within this video kzfaq.info/get/bejne/n8VyjcilptmRdnU.html. Mam ila sir than.

@nagasankar2566 5 ай бұрын

Thank you Mam...You are very good in teaching such of these complicated topics & problems...

@because2022 5 ай бұрын

It's my pleasure. Mam ila sir than.. Share with your friends too.

@PKavin-kz8ob Жыл бұрын

Sir I NEED THEORY OF COMPUTATION FOR FIVE UNITE SIR SO PLSE BUT FAST SIR

@because2022 Жыл бұрын

Wl continue by this weekend.

@mariesuja Ай бұрын

sir ,could you please tell important questions for machine learning unit 4 sir

@because2022 Ай бұрын

Alrdy the in playlist startng..

@sonaashraq Жыл бұрын

Sir ..intha question ku nenga edutha example + theory ah explain panalam tahne ...atu potum la

@because2022 Жыл бұрын

Yes Sona.

@AjithKumarA-n5n 14 күн бұрын

gradient descent explanation plss

@because2022 14 күн бұрын

In this I wud have explained kzfaq.info/get/bejne/n8VyjcilptmRdnU.html

@v12v123 Ай бұрын

1)Elaborate the process of training hidden layers by ReLU in deep networks. 2)Briefly explain hints and the different ways it can be used. 3)Assume an image has pixel size 240 x 180. Elaborate how K means clustering can be used to achieve lossy data compression of that image. How to answer this questions sir?

@because2022 Ай бұрын

For first question it will remain the same, only activation function will be different at layers and similarly the derivative might vary due to different activation function. I am not sure what they mean in second question. I guess applications of Relu. For 3rd question, you can refer www.geeksforgeeks.org/image-compression-using-k-means-clustering/

@v12v123 Ай бұрын

@@because2022 thank you sir☺️

@v12v123 Ай бұрын

@@because2022 sir but u teach training neural network with gradient descent and stochastic gradient descent In this way how can I Use ReLU activation function?

@because2022 Ай бұрын

@@v12v123 I would have derived equation using sigmoid function right? You need to replace it with Relu. Try once and let me know if you dont get it.

@v12v123 Ай бұрын

@@because2022 ok sir is aiml is tooo tough?

@ChandraKishore-mx7ky 17 күн бұрын

14:40 In taking derivation, only we have to consider numerator ah sir? In first step.. Der of out o is not considered and in second step der of net o & in third step der of w2 were not told by you? Kindly explain sir!

@because2022 17 күн бұрын

We should not take numerator and denominator separately. Its basically combined term only we are finding.

@ChandraKishore-mx7ky 17 күн бұрын

@@because2022 Ok sir thanks for the information. Will you please explain in an video about DISCRIMINANT FUNCTION in Linear classification models in Unit 3 sir.

@because2022 17 күн бұрын

@@ChandraKishore-mx7ky Wont be possible within today. You can refer book for time being. Sorry

@ChandraKishore-mx7ky 17 күн бұрын

@@because2022 Ok sir👍

@sivasakthi56460 17 күн бұрын

Sir why we use 1/2 instead of 1/n Rendu weights use pannathunala n kku pathila 2 use pannirukkomaa sir

@because2022 17 күн бұрын

Thats for calculation simplification only. You can use 1/n also

@sivasakthi56460 17 күн бұрын

@@because2022 ok sir if I use 1/n[y-ycap]^2 after differentiating how it will become sir n is constant so it will become -2/n[y-ycap] ithu correct aa sir

@because2022 17 күн бұрын

@@sivasakthi56460 Yes 2 wont get cancelled basically. Rest will remain same

@sivasakthi56460 17 күн бұрын

@@because2022 sir after that what value should I subtitute to n

@collegepasanga2.022 Жыл бұрын

Sir error propagation problem part c la que na kepangala sir

@because2022 Жыл бұрын

May be kekalam or part b la kuda kekalam.

@d.saravanan9263 Жыл бұрын

Sir why bias value add pannala 1:48

@because2022 Жыл бұрын

For simplicity. U can add it

@hariharasudhan.b9615 Жыл бұрын

Sir ai&ml ku note kitaikuma pls

@because2022 Жыл бұрын

I dont have separate notes. But you can use ppt as short notes.

@Movie_shorts_365days Жыл бұрын

Sir 1st example mattun eludhana podhuma sir

@because2022 Жыл бұрын

Write whichever is easy for you.

@nishanthm2058 Жыл бұрын

CS3491 Artificial intelligence machine learning lab manual download link send me

@because2022 Жыл бұрын

I dont have Nishanth. Try to ask some other college students.

@suryaprakashj9411 17 күн бұрын

there is no such ppt in that drive link sir, please mention the name of this ppt sir

@because2022 17 күн бұрын

drive.google.com/file/d/1B6GjP-FUJn8J2oRnNmCivr-eWt3JwNh-/view?usp=drive_link

@suryaprakashj9411 16 күн бұрын

@@because2022 thanks sir

@ayishwaryac9724 Жыл бұрын

sir put soon aiml unit 5 videos

@because2022 Жыл бұрын

Sure can you please send local author pdf copy of 5th unit to my mail venkat.kvhapp@gmail.com if you have.

@memeboys2268 Жыл бұрын

Sir numericals kepangala romba kestama irukuthu sir 😢

@because2022 Жыл бұрын

Mostly theory a iruka vaipu iruku. kzfaq.info/get/bejne/eJeqisx9usvRo58.html

@sivasakthi56460 23 күн бұрын

Sir intha example ye exam la full aa complete panni eluthathukku time irukkumaa sir eppadi eluthurathu sir

@because2022 23 күн бұрын

Just practise it few times. Then you will get good speed.

@sivasakthi56460 23 күн бұрын

@@because2022 sir if I write all the itrations it will take more than 30 minutes instead writing itration 1 is enough or not sir?

@because2022 22 күн бұрын

@@sivasakthi56460 Then you can write 1 iteration or 2 iteration

@sivasakthi56460 22 күн бұрын

@@because2022 ok sir

@prince4745 Ай бұрын

Sir Neural networks la back propagation iruku antha question ku neenga explain pana example ah potu key points and algorithm eluthuna pothuma

@because2022 Ай бұрын

Yes

@dinesh6173 18 күн бұрын

Bias value um updatioon um illaye sir

@because2022 18 күн бұрын

In actual bias also would be updated.

@dinesh6173 18 күн бұрын

@@because2022 weight ku use panra same formula va sir

@because2022 18 күн бұрын

@@dinesh6173 Yes only derivative will change

@dinesh6173 18 күн бұрын

@@because2022 ok sir thank you

@because2022 17 күн бұрын

@@dinesh6173 Formula will remain almost same, but will be in terms of bias.