Definitely saving this playlist! Thank you so much, your videos answered a lot of questions I had from my Biostats class, and you explained it so clearly.
Wonderful series of Lectures. .. please upload more topics on Survival Analysis. Also, here is my request - kindly upload a series of lectures on reliability theory, if possible. Thank you.
Just in case someone else is a mathematican here: The author actually tells in the next video, taht the hazard is actually the derivative of the survival function with respect to time. That confirms my believe that the probability in the definition of hazard should be divided by the delta (and taking the limit for delta going to zero). Even without the limit, with the division by delta the definition makes much more sense as it is by far not as depndent on the value of delta as long as it is small enough so that the survival function can be approximated by a line. Without the divisin, taking delta twice as large makes the "hazard" about twice as high and it makes it uselss. But I understand that most of the audiance are not mathematicians comming here just for statistics. Explaining this in a more detail could result in more confusion than clarification.
If the event that I want to analyze is the time to improvement the condition of a patients (in this case based on their hospitalization during xx days), then what is the definition of survival function?
Is "the probability of patient's condition to improve beyond time t"? If so, that surv definition doesn't match with the characteristics of the survival function curve which should be monotonically decreasing. (bcs the real data show that the longer the patients are stay,then the more they will heal)
I think you are looking at "the probability that the disease can survive from treatments after t days". Also, you can interpret this survival probability as the "effectiveness of treatment", the longer t the smaller of S(t) and less effective of the treatment (or worse of the patients initial condition). It is depending on whether you are analyzing risk factors or analyzing treatments.
![Hazard and Survival Functions - [Survival Analysis 5/8]](/img/1.gif)
100 times better than the class I joined!
Wonderful lecture!!! You filled the void of good lectures on this topic....May Almighty bless you.
Definitely saving this playlist! Thank you so much, your videos answered a lot of questions I had from my Biostats class, and you explained it so clearly.
You are still saving lives. Thank you
3.59 lightbulb. You are an amazing teacher, exam soon.. THANKS!
You've made the concept so easy to understand.
Thanks :)
What an excellent teacher
HOW AMAZING THE VIDEO IS!! BETTER THAN MY TEACHER SAID!
Thank you for the very easy to understand lecture!
So thankful for your contribution.
You explain this so clearly! Great.
This really helps!! Thank you!
Amazing didactic, very well explained. Thank you very much!
Amazing Work
Thank you so much, you make this easy to understand
Thank you! Very helpful
Very clear, thank you
Wonderful series of Lectures. .. please upload more topics on Survival Analysis. Also, here is my request - kindly upload a series of lectures on reliability theory, if possible. Thank you.
Just in case someone else is a mathematican here: The author actually tells in the next video, taht the hazard is actually the derivative of the survival function with respect to time. That confirms my believe that the probability in the definition of hazard should be divided by the delta (and taking the limit for delta going to zero). Even without the limit, with the division by delta the definition makes much more sense as it is by far not as depndent on the value of delta as long as it is small enough so that the survival function can be approximated by a line. Without the divisin, taking delta twice as large makes the "hazard" about twice as high and it makes it uselss. But I understand that most of the audiance are not mathematicians comming here just for statistics. Explaining this in a more detail could result in more confusion than clarification.
Thank you sir
Thank you very much for this!!!!!
Thanks!
Thank you for this!
Excellent, thank you very much
This is amazing!
Thanks
If the event that I want to analyze is the time to improvement the condition of a patients (in this case based on their hospitalization during xx days), then what is the definition of survival function?
Is "the probability of patient's condition to improve beyond time t"?
If so, that surv definition doesn't match with the characteristics of the survival function curve which should be monotonically decreasing. (bcs the real data show that the longer the patients are stay,then the more they will heal)
I think you are looking at "the probability that the disease can survive from treatments after t days". Also, you can interpret this survival probability as the "effectiveness of treatment", the longer t the smaller of S(t) and less effective of the treatment (or worse of the patients initial condition). It is depending on whether you are analyzing risk factors or analyzing treatments.
good explanation