Yes, when there is no overlap and the two curves are perfectly separated, the AUC is 100%. I guess the first ROC (at 7:22) has implicit extending curves that are not drawn.
I dont mean to be offtopic but does someone know a method to get back into an Instagram account..? I was dumb lost my password. I appreciate any help you can offer me!
@Gideon Dariel I really appreciate your reply. I got to the site through google and im trying it out atm. Seems to take a while so I will reply here later with my results.
I really was struggling to understand this material upon learning it in my Psych class, but after watching this video I completely understand it now. Thank you for such a clear and simple explanation! Subscribed :)
Excellent Explanation! These principles affect EVERY field: satellite comms, or encoding symbols in magnetic storage mediums like HDDs, = "inter symbol interference"; Medical tests = you covered it; statistics = "inference testing"; Microscopy and remote sensing = spectral resolution (rayleigh criterion).....everywhere in science!
Good explanation, Rahul. I've recommended this to students in my S&P course before, but have stopped due to the questions regarding what the distributions represent. As described early on, your x-axis represents levels of a particular test. What tends to be challenging for students is entertaining that a single test for a disease can have a range of values; they tend to consider it as you either do or do not have a disease. COVID tests may be a good example to consider to clear up any confusion. Older tests have been used to detect if you have the virus or not. The histograms in the video would not represent that kind of test. The newer tests are looking to determine if someone may be immune to the disease based on antibody levels; lower levels indicate more vulnerability. That type of test would work with your examples--the blue representing populations that are still vulnerable, the green representing populations that are immune. This is still histograms of populations, but the ranges make more sense. What's yet to be determined is what anti-body level is appropriate to confidently state that someone is immune. Given how little we know about the disease, you could extend this example to include more overlap (a lower d') and a shift of the cutoff to the right to encourage specificity--the public health benefit is not to determine whether an individual has antibodies, but whether they have enough antibodies; only the individuals who have the highest levels of antibodies would be confidently designated as immune (though still contagious).
wowowowowowowwo you saved my life!!!!!!!!!!! i thought i was naturally unable to comprehend this concept when my prof was teaching us in his way, but after your video, it seems that i'm not too stupid :D
Thank you for this! I was having a hard time understanding ROC's with the book, but this made it so much easier to understand! Thank you so much! Also loved the radar stuff at the end! :)
This video explained ROC and AUC 1000x better than my professor did. Thank you!!!
Excuse me, when there is no overlap between the two curves, The ROC should be 100%, not 90%.
Yes, when there is no overlap and the two curves are perfectly separated, the AUC is 100%. I guess the first ROC (at 7:22) has implicit extending curves that are not drawn.
I dont mean to be offtopic but does someone know a method to get back into an Instagram account..?
I was dumb lost my password. I appreciate any help you can offer me!
@Holden Rey instablaster :)
@Gideon Dariel I really appreciate your reply. I got to the site through google and im trying it out atm.
Seems to take a while so I will reply here later with my results.
May I say that this was SO clear and clinically relevant. THANK YOU A TON from a doc who needs to understand big picture fast!
I really was struggling to understand this material upon learning it in my Psych class, but after watching this video I completely understand it now. Thank you for such a clear and simple explanation! Subscribed :)
Thank you sir for solving the 'mystery' of the 1-Specificity. It was truly confusing but now crystal clear.
You have made this very confusing topic very easy and crystal clear. Thank you Sir.
Excellent Explanation! These principles affect EVERY field: satellite comms, or encoding symbols in magnetic storage mediums like HDDs, = "inter symbol interference"; Medical tests = you covered it; statistics = "inference testing"; Microscopy and remote sensing = spectral resolution (rayleigh criterion).....everywhere in science!
Brilliant. After years of studying these concerts I finally get it!
Loved the way you brought history and WWII into med stats!!!
You are a good teacher, the best explanation of ROC and AUC I have come across yet!
wow.. this is by far the best explanation ive heard about the ROC curve.. thank u so much sir.. God bless u.. :)
Excellent explanation of ROC and AUC. The icing is how ROC got its name!
Excellent.. I 'll never forget these concepts and excellent presentation..
l have been trying for a while to understand this concept..best explanation..thank you and God bless
Damn!, why didn't I find this video first? You got a new subscriber with this explanation. On point, clear, no confusion!
Wonderful explanation! This type of curves are also applied to check how good a model of species distribution discriminate
This was the best explanation I have found so far!!!!!
Excellent! Great explanation of ROC that I've seen!
Thank you so much, great explanation!
Yaaa ofcourse!
Better explanation👍
Brilliant! it is just brilliant. No other words can describe my joy after cracking this concept.
Great congratulations and thank you for sharing
Great job. I really liked the use of colors to explain. It made it very easy to understand. Thanks
BEST EXPLANATION ! THANK YOU!
could never be better,,,great thanks
Excellent explanation, thank you!
Truly one of the best videos on the subject!
Excellent, very clear and concise!
Beautiful explanation!!
Thank you.
Great video. Cleared a lot of confusion. Thank you
This video explained ROC and AUC better than that any of what i have seen in China.
Good explanation, Rahul. I've recommended this to students in my S&P course before, but have stopped due to the questions regarding what the distributions represent. As described early on, your x-axis represents levels of a particular test. What tends to be challenging for students is entertaining that a single test for a disease can have a range of values; they tend to consider it as you either do or do not have a disease. COVID tests may be a good example to consider to clear up any confusion. Older tests have been used to detect if you have the virus or not. The histograms in the video would not represent that kind of test. The newer tests are looking to determine if someone may be immune to the disease based on antibody levels; lower levels indicate more vulnerability. That type of test would work with your examples--the blue representing populations that are still vulnerable, the green representing populations that are immune. This is still histograms of populations, but the ranges make more sense. What's yet to be determined is what anti-body level is appropriate to confidently state that someone is immune. Given how little we know about the disease, you could extend this example to include more overlap (a lower d') and a shift of the cutoff to the right to encourage specificity--the public health benefit is not to determine whether an individual has antibodies, but whether they have enough antibodies; only the individuals who have the highest levels of antibodies would be confidently designated as immune (though still contagious).
Damn... That was excellent... Thank u for making it clear ..
fantastic video! Pace and explanation exemplary
Thank you for the clear explanation. Save me lots of time on reading but still not understanding!
Thank you so much, your explanation on this topic is very easy to digest yet very informative.
Excellent explanation , u are a great teacher , thank you ❤️
Wow, this is the best video explaining ROC and AUC
fantastic. great clarity
wowowowowowowwo you saved my life!!!!!!!!!!! i thought i was naturally unable to comprehend this concept when my prof was teaching us in his way, but after your video, it seems that i'm not too stupid :D
what a vivid explanation !!
Much more useful
Love from 🇮🇳 India
Thorough and clear. Thanks.
great visual explanation!
Here comes another subscriber..you explained it so well!
I went straight to subscribe button after watching your vedio....
Thanks, I really learned a lot from this video, but more so from the previous video on sensitivity and specificity.
very very efficient and understandable ,thank you
what a fantastic video- thank you
Super explanation. Loved it. Would like to see more of your videos now
Thanks, very nicely explained.
excellent review!Thank you soooo much
superb ...love this video
great job Rahul. Thank you very much! Keep on!
Thank you! Very well explained!
Extremely well put together ! Thanks a million .
Thank You so much sir.
Excellent RUclips explanation of ROC and AUC
Thank you for the presentation! Really like the example!
Excellent presentation!Thank you very much!
Great explanation! Thanks!
excellent tutorial. many thanks
Beautifully explained. Thank you.
Excellent video, congrats!!!
Awesome. Really helpful. Thank you
brilliant explanation
Dude you totally saved my midterm
Very good explanation.
very well explained and illustrated - perfect! Thank you very very much!
Awesome video. Thank you so much for all your work!
Thank u very much.... Much appreciated... 👍👍
Excellent video
Great explanation sir! Thank you for your time explaining this.
very well explained.... Thank rahul.
Amazing video! Great explanation. Cool and calm voice!
Thank you so much for your excellent explanations!!!
Awesome video... Thank you for such a clear explanation
Nice history lesson!
You really explained ROC well. Thank you!
Outstandingly clear!
This is great ..very clear and communicative..thanks
That was a real piece of art :) Thanks
Great job! Thank you so much!!!
very concise, thank you! cool WWII info
Thank you for this! I was having a hard time understanding ROC's with the book, but this made it so much easier to understand! Thank you so much!
Also loved the radar stuff at the end! :)
the best one I have seen!
Brilliant! Very helpful and informative..
Excellent Video
Really awesome explanation man.... So clear and cool... keep it up
Very nice video, thank you!
Thank you so much! This was very helpful
Very well explained.
That's what I call an explanation. Thank you!
Great work!
Thank you! Really good video.
Nicely done, sir!
you are a genius. Thanks
Perfect explanation. thx
Really great stuff!
You're like the Brandon Foltz of Epidemiology! Love the videos, very well explained and easy to grasp in a couple of minutes. Bravo
wow, thanks. one of the best explanations i've seen!
Terrific, clear and interesting!!!! Thank you!!!