Volume : V, Issue : VIII, August - 2015

MODIFIED AUTO REGRESSION MODEL FOR RACE OF ADULTS

S. C. Thasleema, B. Sarojamma

Abstract :

Auto regressive model is relation between present time series variable and past time series values. Generally regression gives relation between dependent variable and independent variables. In time series a univariate variable relates chronologically. In this paper two modified auto regressive models by incorporating time to first modified auto regressive model and second modified autoregression model is incorporated by adding time and also dividing factorials to past time series values. Root mean square error criteria are used to choose best among three models i.e. auto regressive model, modified autoregressive model–1 and modified auto regressive model–2. The data“percent of people 25 years and over who have completed high school or college, by race, Hispanic origin and sex, for the period, of 23years i.e., 1990–2014(non institutionalized population)”. Among the fitted 15 models, 5 for auto regressive, 5 for modified auto regressive model–1 and 5 for auto regressive model–2 for orders 1, 2, 3,4 and 5. Among three auto regressive models, modified auto regressive model–2 is the best for the given data.Among 15 models, 5th order modified auto regressive model–2 is the best

Keywords :

Article: Download PDF   DOI : 10.36106/ijar  

Cite This Article:

S.C. Thasleema, B. Sarojamma Modified Auto Regression Model for Race of Adults Indian Journal of Applied Research, Vol.5, Issue : 8 August 2015

Number of Downloads : 589

References :