mean+standard deviation+skewness+kurtosis

WEIBO

New Member
#1
hi

I want to order some variables according to (mean+standard deviation+skewness+kurtosis) or some new rule considering mean value, standard deviation, skewness and kurtosis simultaneously. Is there any similar sorting rule like this? Is this equation used in other research ares?

thank you.
 

Dason

Ambassador to the humans
#2
It's not really clear to me what you want to do. I think the more important question though is what you're *actually* trying to do. What is it that you want to order the variables by and why?
 

WEIBO

New Member
#3
It's not really clear to me what you want to do. I think the more important question though is what you're *actually* trying to do. What is it that you want to order the variables by and why?

The variables are data sets which include several different numbers. I want to order these sets according to their mean, standard deviation, skewness and kurtosis. Is there any similar method used for some research area? or how to determine the weights of each parameter, such as a*mean+b*std+c*ske+d*kur?

thank you
 

Dason

Ambassador to the humans
#4
That didn't help at all. WHY do you want to do this. What is your actual purpose here. I'm assuming you're not just ordering them and then saying "Good job me! Now I'm all done". Why do they need to be ordered. What is your actual goal.
 

hlsmith

Not a robit
#5
WEIBO objective is to burden Dasonbot processing power. Objective achieved. Repeat objective indefinitely until objective failure, then engage machine learning and execute new burdening actions.
 

WEIBO

New Member
#6
That didn't help at all. WHY do you want to do this. What is your actual purpose here. I'm assuming you're not just ordering them and then saying "Good job me! Now I'm all done". Why do they need to be ordered. What is your actual goal.
The goal is to order them according to their mean value and variation. That is the set with large mean value and large variation is ordered first.
 

maartenbuis

TS Contributor
#8
Say you have two variables A and B. The mean of variable A is larger than the mean of variable B, but the variance in variable A is smaller than the variance in variable B. How would you order them? There is no magic solution, that says "ofcourse variable B should be before variable A". It depends on how you weight the mean and the variance. The weight represents how important you find these characteristics. Neither I nor Dason are you (I think), so we cannot decide for you what you should consider important. We might have been able to help you if you told us why you wanted to order those. That might help us derive the weights, or at least give you a suggestion on how to get started or limit the number of possible weights. Since you are unwilling to tell us why you want to order the variable, there is literally nothing we can do. The problem you possed contained insufficient information to come even close to an answer.
 

hlsmith

Not a robit
#9
Say like a meta-analysis, can you inverse variance weight them. Come on you all, you can't let WEIBO win, or leave without an adequate solution.