Regression Definite Integrals Using Only Observations with Positive σ-Contribution
✍ Scribed by Dr. C. A. Glasbey
- Publisher
- John Wiley and Sons
- Year
- 2008
- Tongue
- English
- Weight
- 420 KB
- Volume
- 32
- Category
- Article
- ISSN
- 0323-3847
No coin nor oath required. For personal study only.
✦ Synopsis
JHctfy
-An algorithiii is proposed for estimating the value of a definite integral in a regression model, subject to the constraint that increasing the value of any observation does not decrease the estimated value of the integral. It is shown t h a t the estimator is consistent and asymptotically unbiased. It also has minimum variance in the special case of simple linear regremion. I n estimating the average illilk yield of cows in 100-day periods, using monthly observations and various nonlinear regreesion functions, the method produces estimates with loner mean square error than either least squares regression or simple averaging.