In: Statistics and Probability
suppose the 2000 US census is collected for every person in the US and it allows researchers to calculate summary quantities like the percent of the US population in a specified age range. What is the mean squared error of these summary quantities calculated from the census?
ANSWER::
The mean square error is the arithmetic mean of squares of errors and is calculated in the following steps:
*. According to 2000 census, suppose the US population is divided into k number of specified age ranges (classes).
*. For each range, the actual percentage of population is calculated as
(Population size in the specified range/Total population)x100
*. Suppose the above actual percentages are assumed to follow any one of theoretical probability distribution suitably for future prediction
*. Then find the expected percentages as
(Total Population)x(Probability function value) for each specified age range
*. Now, error is the difference between actual percentage(a) and expected percentage(e), particularly, for an ith specified range, the error is
Ei = ai-ei
*. The above Ei s and Ei 2 s are calculated for all k specified age ranges.
*. Finally, the Mean Square Error is calculated as
MSE = [E1 2 +E2 2 +E3 2 +…+Ek 2]/k
*. If MSE is very least and nearer to 0, then the assumed theoretical probability distribution used to find expected percentages may be more suitable for the actual percentages.
NOTE:: I HOPE YOUR HAPPY WITH MY ANSWER....***PLEASE SUPPORT ME WITH YOUR RATING...
***PLEASE GIVE ME "LIKE"...ITS VERY IMPORTANT FOR ME NOW....PLEASE SUPPORT ME ....THANK YOU