I´m writing you from the future, 4 years after have done this incredible video, and I have to say "thank you": this is the only video in TH-cam (trust me) that really explain how to do a RSM example in R. Incredible useful. No words
Great explanation about how to optimize one response. Actually, that dataset contains two more response variables, and I was wondering, how could we optimize the three responses at once?
You need to use a multi-response optimization technique such as the desirability function, genetic algorithms etc. First, consider optimizing one response at-a-time using factorial designs and response surface methodology. You should obtain mathematical models (one mathematical model for each response = 3 mathematical models in total). Finally, apply the desirability function and initialize the objective (either to maximize, minimize or arrive at a target value). The desirability function will run the three mathematical models simultaneously and generate a set of process parameter settings that yield the best combinations for the three responses. Lastly conduct experiments to validate the optimum process parameter settings obtained from the desirability function.
@@DennisMutuma I have a system with two steps that affect the yield. This is an anaerobic digestion process that has a 1st and 2nd digester. Each digester has 3 variables that affects its functionality (temperature, retention time and water content) combined with the feedstock (bio waste), totalling 7 independent variables that affect determine the yield. The experiment has 365 observations. What will be the best optimization method using RSM to result to a greater yield?
I´m writing you from the future, 4 years after have done this incredible video, and I have to say "thank you": this is the only video in TH-cam (trust me) that really explain how to do a RSM example in R. Incredible useful. No words
Seriously. It's the only one... STILL. Great content.
This is very helpful! Thanks!
Thank you so much, your lectures are gems
Great video. I highly appreciate the information.
Very nice Explained. Thank You!!
It was very useful to me. Thanks
Thank you very much for doing this!
Great explanation about how to optimize one response. Actually, that dataset contains two more response variables, and I was wondering, how could we optimize the three responses at once?
There is only one response - Yield. There are three input variables.
You need to use a multi-response optimization technique such as the desirability function, genetic algorithms etc. First, consider optimizing one response at-a-time using factorial designs and response surface methodology. You should obtain mathematical models (one mathematical model for each response = 3 mathematical models in total). Finally, apply the desirability function and initialize the objective (either to maximize, minimize or arrive at a target value). The desirability function will run the three mathematical models simultaneously and generate a set of process parameter settings that yield the best combinations for the three responses. Lastly conduct experiments to validate the optimum process parameter settings obtained from the desirability function.
@@DennisMutuma I have a system with two steps that affect the yield. This is an anaerobic digestion process that has a 1st and 2nd digester. Each digester has 3 variables that affects its functionality (temperature, retention time and water content) combined with the feedstock (bio waste), totalling 7 independent variables that affect determine the yield. The experiment has 365 observations. What will be the best optimization method using RSM to result to a greater yield?
that's very helpful, thanks for making this vedio
So helpful.
The level of significance used was 0 for the entire experimental design?
Could we get the R code?
You can the R code for this and all the other lectures here: www.lithoguru.com/scientist/statistics/course.html
Good
no offense. But its a little complicated for me. I need more back ground information to understand DOE i guess.