Does anyone know why do we have k>1 in the formula of Chevychev's inequality. I know this is for any distribution, not limited to normal distribution. Sign of K should not make any difference as we have K^2 in the formula. I just want to make sure I am not missing anything here.作者: Wasteoftime 时间: 2011-7-13 16:29
THe formula is 1- 1/(k^2).
If k=1, then the result is 0 -> doesnot make sense作者: dcfox83 时间: 2011-7-13 16:29
also if k is between 0 and 1, when you square it, it gets smaller. Whereas a number above 1 obviously gets bigger.....作者: rkapoor 时间: 2011-7-13 16:29
I know but how about K<0. Why is that not included? Any specific reason or just to avoid these limits stated by you guys..作者: hariRaj 时间: 2011-7-13 16:29
It's ok to have K <0 bcos we square K in the formula. Positive or negative both yield the same result.作者: eagles_dare13 时间: 2011-7-13 16:29
K is the number of stds away from the mean right? so it cant be negitive because you would use absolute value?
