Mod-01 Lec-29 Support Vector Machine

[@hiteshkaushik2085]

Cant find a better lecture on SVM than this. World class. Better than MIT and stanford anyday.

[@nilarunm]

I have listened to SVM e-classes from various sources, including Stanford,but this is the most explanatory and understandable, Sir teaches from realization, not from comprehension

[@varunmanjunath6204]

exactly.India is underestimated in terms of education

[@karthik-ex4dm]

Lectures from Stanford, MIT etc etc cannot come anywhere near this explanation!!!! Not only in ML, But in most of the topics #BestInTheWorld

[@sanjaykumarbahuguna4030]

I have done lot of search to understand SVM & I found you are the best. Thanks sir.

[@vichitrakumar4452]

i have tried understanding it from so many sources, this one made me understand, far better than any other source on the internet, Thanks!!

[@imalive404]

kzbin.info/www/bejne/lYHamZyNra1-btE

[@santhoshnarayanan343]

So far the best best lecture making understand each and every aspects of Support Vector Machine theoretically.

[@grafdp]

Very well explained, I don't understand why this hasn't much more clicks. A true saviour.

[@siddharthabaral4271]

Why cant other videos be this simple ? This is just brilliant.

[@PankajKumar-tr7bg]

Gr8 resource to watch far better then MIt Stanford.

[@adbcify]

This is far better than MIT or Stanford... seriously!

[@imalive404]

This video is good. but you might want to check this kzbin.info/www/bejne/lYHamZyNra1-btE

[@generationwolves]

@@imalive404 Or this: kzbin.info/www/bejne/n5yadmqoZ8Zknpo

[@geethujoyanand]

The best lecture on SVM.thank you Sir.

[@binnysplaytime4640]

Excellent class Sir

[@williamstephenjones3863]

Brilliant!

[@sandhyapotadar358]

Awesome explanation sir. Thank you

[@MohammedHarisSumair]

This lecture was very easy to comprehend and was also enjoyable.

[@meens1986]

Awesome explanation..thanks a ton for making it so clear

[@vishaldhawan420]

SVM starts from 5:13

[@shivang7989]

Thanks

[@mazharulislam7619]

Excellent Sir

[@aniruddhg5641]

Wonderfully explained Sir. Thank you very much.

[@Code-09]

Best lecture. Thank you very much.

[@gauravsharma-mi2er]

One of the best video.

[@bibhuprasadsahu5595]

good explanation of SVM.

[@JourneyMindMap]

Thank you, difficult to find the clear explaination like this

[@shivendrapanicker8233]

Awesome lecture.

[@adityajoshi5]

Intro music sounded like background score in romantic scene of a 90's movie.

[@TonySahoo92]

To me it sounds more like a 'thriller'

[@abhishek007123]

😂😂😂😂😂😂😂

[@bishwanathlalkarn8203]

thank you...it's awesome lecture... it's really help me in my thesis work....do you have any lecture video on Fisher Kernel.....

[@RKDataAnalytics]

Appreciate the efforts, Thanks

[@AnilSingh-vx4zs]

watch from 27:36

[@insoucyant]

Wonderful class. Thank You.

[@adewolekayode6148]

Thank you so much sir for this lesson. Sir, please can you present a tutorial on TSVM? PLS

[@rakhsupddr1034]

Thank you sir , the lecture was really good. I have a doubt though : *) We had to maximize the distance between 2 support vectors (1 +ve and other -ve), that's why we minimized 'W' (because it was inversely proportional to the distance), but at 46:46 (time of the video) you told to MAXimize the 'Lagrangian' exp (which is directly proportional to 'W') to find the 'alpha_sub_i' 's. I didn't find any reason for this Maximization in particular (As it has already been MINimized to substitute for 'b' and 'W').

[@kamaluddindimas1661]

its a dual problem bro.. you can check in on calculus subject

[@eswan7638]

Excellent

[@SumitChauhan009]

How did you write the value of b margin as that ?

[@sanjeevdhewa9670]

for svm start from 5:18

[@bicepjai]

good one

[@shailendra9292]

At 5:20 Support vector machine

[@gou-goutham]

Thank you sir....

[@harshinisewani5095]

at 44.35 the summation having alpha and y of i why it didnt gets cancel.

[@nishugarg7466]

thank you sir

[@PSNAcademy]

At 43:29, the 2nd term is getting cancelled due to sum(aplha[i].y[i])=0. Then why not the 3rd term term?

[@kamaluddindimas1661]

the 2nd term can be cancelled because b is an constant, so the the sum now consist of alpha i and y i. but in the 3rd term, the sum consisit of alpha, yi and w...

[@nikhilashodariya]

Sir, can you please tell why at 44.59 (time of the video) for W.W two different subscript was used why can't we use alpha_i*alpha_i * y_i*y_i * x_i*x_i instead of alpha_i*alpha_j * y_i*y_j * x_i*x_j

[@prof.manjeetsinghjcboseust9034]

it is for more clarity in expression

[@dilipparmar7715]

Thanku you !! Appreciates !!

[@umarfarook7773]

thank you sir .how can we find the value of alpha(i)?

[@kamaluddindimas1661]

when you expand the maximize of langrangian dual... you can get new equation, and take derivation of alpha.. and you can compute the alpha

[@aishwaryanarkar2954]

thank yuuuu so much :)

[@suqbah1819]

how did he get the equation for b?

[@rakhsupddr1034]

You can find that out once you know the support vector(s). According to THIS video we find all the 'alpha_sub_i' and we know all the 'X_sub_i' as well as corresponding 'Y_sub_i' and hence we find the 'W' which is nothing but the summation of 'alpha' 'X' and 'Y' . After this for any support vector we know 'WX + b = +1/-1' (so we know 'W' ,'X' and can find 'b'.

[@arnimachaurasia7]

thank you sir.... :-)

[@venukarthikboddu7031]

Can any one explain how this SVM can be implemented practically like by taking a dataset containing reviews and building this SVM model for classifying reviews into positive and negative classes???????

[@sanketdamane2642]

Sentdex have good videos on youtube for svm using python

[@K_SE__VishalRoyRoy]

Can any one who could guide me how to make a roadmap to learn machine learning

[@hassuunna]

1.5X speed

[@sidhantthole4185]

46:45 it should be minimization

[@shubhamvashisth9518]

No, it will be maximize only

[@harshinisewani5095]

@@shubhamvashisth9518 why it will be maximized? please can you explain

[@shubhamvashisth9518]

Because it's the property of the Lagrangian method to maximize factors.

[@vishalsahu4233]

u bores students a lot! tomr is my exam and I am wasting time in this.

[@jatinlalwani9459]

Disappointed :(

[@sarumathiykulasekaran8841]

Svm explanation starts from 5:38

[@shrikantgangamwar]

Maki aakha teri

[@wasimbashir8279]

Thank you sir

[@patelroshni8548]

thank you sir....

[@hiteshsomani4462]

Thank you sir...