Obviously, these videos are super helpful. JFYI, these videos are supplemental study material for an A.I. study group that I'm mentoring. Your slides are so nicely laid out. I suspect that you really wanted to be an industrial designer but your parents made you go into medicine. :) Many thanks!
Hi, thank you so much for the amazing explanation. I want to ask regarding the last point that you taught. lets say, if the question asks us to choose the curve with highest specificity, and they have given different curves with different AUCs. Will we still choose the left most point curve? like if the point which is the left most among all the curves has its AUC lower than the curve whose specificity point is relatively to the right but its AUC is more. which curve will have highest specificity? I got it complicated, hope you get my question!
Great question! I understand where it feels complicated. My understanding is that every line represents a test and the different points along that line represent different cutoffs. For such a question to be valid, they would have to focus on ONE point and DIRECTLY compare it to another point. When it comes to AUC, it is just a simple way to tell if one test if better performing than the other. I doubt a question would show you two lines and then ask you which test is more specific as it truly depends on which cutoffs they specify. Such a question where the curves are intermixed such that specific cutoffs can be found more on the left with the overall AUC less than another test is in my opinion a very unfair question that doesn’t target any true objectives of the concept. I hope that made sense. If anyone else can shed light on this, I would appreciate seeing your thoughts.
Hmm, that’s a neat question. If I recall correctly, the AUC does not tell you anything about the sample size. The ROC curve in general is a visual depiction of the different cutoffs that could be used as well as their associated sensitivities and specificities. These are good measures to highlight how well the test can rule out or rule in disease. I don’t believe they say anything about the sample size.
![VICTOR Denmark Open 2024 | An Se Young (KOR) [1] vs. Gregoria Mariska Tunjung (INA) [5] | SF](http://i.ytimg.com/vi/Y3QEyCih85s/mqdefault.jpg)
![Wiz Khalifa - Khalifa's Home [Official Music Video]](http://i.ytimg.com/vi/wnrLCTe0Nz4/mqdefault.jpg)
I got 2 questions from this curve in my Step 1 exam. Thank you so much for the explanation. It really helped me to answer those questions.
This was the simplest explanation I could find. Thank you so much
You’re most welcome! 🙏🏼
Simplified !
Thank you.
Simple and straight to the point. Thank you
thank you! I 'd been struggling for a while until I found this video :D
Great quick review, like I love it 🤠
Thank you for this short and clear video. am very grateful!
This was the best explanation i could find. Thank you!
Thank you for the kind words!
Thanks! 🎉 Loved this simple explanation. Easy to remember
I just want to say thank you for this video! You just saved me a lot of stress right now.
The simplest and very clear explanation.. Thank you
You’re most welcome! 🙏🏼
Amazing explanation and to the point
Obviously, these videos are super helpful. JFYI, these videos are supplemental study material for an A.I. study group that I'm mentoring. Your slides are so nicely laid out. I suspect that you really wanted to be an industrial designer but your parents made you go into medicine. :)
Many thanks!
Umm… okay.
Dig the video. Good to see more Medutubers unite ! subbed
thnkyou. short and precise
You’re welcome, Fahad! I’m glad you found the video useful.
Well explained!!
Thank you so much.
You’re most welcome!
Thank you so much you made it so much easier for me 🤗
Thank you Dr
wow amaizing!! thanks!!
You’re most welcome!
Great Explanation
Excellent!!
Thank you!
Awesome, thank you!
You’re the one who’s awesome, friend! 👉🏼👉🏼
Thankyou, It is really helpful Sir
You’re most welcome, Paramita! I’m glad you found it helpful.
Very precise and understandable
Nice work .
Thank you!
Thank you so much god bless
really well explained
Thank you, friend! I’m glad you benefitted! 🙏🏼
Much appreciated!
🙏🏼
Brilliant
Nice video
Hi, thank you so much for the amazing explanation. I want to ask regarding the last point that you taught. lets say, if the question asks us to choose the curve with highest specificity, and they have given different curves with different AUCs. Will we still choose the left most point curve? like if the point which is the left most among all the curves has its AUC lower than the curve whose specificity point is relatively to the right but its AUC is more. which curve will have highest specificity? I got it complicated, hope you get my question!
Great question! I understand where it feels complicated. My understanding is that every line represents a test and the different points along that line represent different cutoffs. For such a question to be valid, they would have to focus on ONE point and DIRECTLY compare it to another point.
When it comes to AUC, it is just a simple way to tell if one test if better performing than the other. I doubt a question would show you two lines and then ask you which test is more specific as it truly depends on which cutoffs they specify. Such a question where the curves are intermixed such that specific cutoffs can be found more on the left with the overall AUC less than another test is in my opinion a very unfair question that doesn’t target any true objectives of the concept.
I hope that made sense. If anyone else can shed light on this, I would appreciate seeing your thoughts.
@@Khalemedic It is immensely helpful, thanks so much. It does make sense. Much appreciated!
Hi can we find sample size based on AUC?
Hmm, that’s a neat question. If I recall correctly, the AUC does not tell you anything about the sample size. The ROC curve in general is a visual depiction of the different cutoffs that could be used as well as their associated sensitivities and specificities. These are good measures to highlight how well the test can rule out or rule in disease. I don’t believe they say anything about the sample size.
how are you able to digest such things and understand them well? do you have any approach you can share with us will be very grateful
I do plan on making a “How to Study Productively” video so stay tuned 😉
@@Khalemedic Awesome subbed. i believe many of these videos miss the systematic approach they use on their courses and focus on time management mostly
If the question asked, which line is "most accurate", which line should you pick?
You should select the line with the highest area under its curve.
@@Khalemedic thanks!
MONEY!