Regression and Analysis of Variance
Kishor Shridharbhai Trivedi · 2016
This chapter presents study aspects of statistical relationships between two or more random variables. In a computer system the throughput Y and the degree of multiprogramming X might well be related to each other. The chapter discusses the correlation analysis. A related problem is that of predicting a value of system throughput Y at a given degree of multiprogramming X. The problem is to find a regression curve that describes the dependence of Y on X. Another related problem is that of least-squares curve fitting. The chapter considers regression when the needed parameters of the population distribution are known exactly. In many problems, the important objective may be to detect and estimate trends in monitored parameters over time. The slope of the trend is then estimated to give a measure of the effect of aging on analyzed operating system resources. The chapter describes the Mann-Kendall test for detecting trends and Sen's nonparametric slope estimator.