RSM318H1 Chapter Notes - Chapter 3: Simple Linear Regression, Linear Regression, Standard Deviation
Document Summary
Is there a relationship between dependent and independent variable. How accurately can we predict y given x. How much does each x contribute to y. Need to separate out individual effects of each predictor. How accurately can we estimate the effect of each predictor on response. How much will y increase given increase in x. If not, may be possible to transform predictor or response so that linear regression can still be used. Can use standard errors to compute confidence intervals. Standard errors can be used for hypothesis testing, through t-statistic. Measures how many standard deviations away from zero. Need to assess extent to which model fits data. Average amount sample mean differs from true mean. Residual standard error: estimates the standard deviation of . Tss: amount of variability inherent in the response. Rss: amount of variability left unexplained by regression. R2: proportion of variability in y that can be explained using x.