Spline Regression Models

Series Editor's Introduction v General Introduction 1(5) Polynomial Regression Models 3(1) Spline Knot Locations Known in Advance 3(2) Splines With Unknown Knot Locations 5(1) Unknown Number of Spline Knots 6(1) Introduction to Spline Models 6(10) Interrupted Regression Analysis 7(2) Piecewise Linear Regression 9(4) Cubic Polynomial Regression 13(1) Important Features of Spline Models 14(2) Splines With Known Knot Locations 16(17) Linear Spline Regression Models 16(7) Quadratic and Higher Order Spline Regression Models 23(1) Hybrid Spline Regression Models 24(1) Model Comparison Issues 25(2) Model Selection Criteria 27(4) Polynomial Regression and Perfect Multicollinearity 31(1) F Statistics and t Statistics 31(1) Autocorrelation and the Durbin-Watson Statistic 32(1) Splines With Unknown Knot Locations 33(16) Transforming Discrete Response Into Continuous Measure 33(2) Interrupted Regression Analysis 35(1) Adjusting Intercepts Only 36(2) Adjusting Intercepts and Slopes 38(1) Splines With Known Knot Locations 39(4) Unknown Spline Knot Location Estimation 43(3) Quadratic Spline With Unknown Knot Locations 46(2) The Wald Test 48(1) Model Selection Conclusion 48(1) Splines With an Unknown Number of Knots 49(12) Stepwise Regression as a Powerful Nonparametric Method 49(3) Determining the Number, Location, and Degree of the Spline Knots 52(1) Smooth Splines for Long-Term Investing 53(2) Moderately Sensitive Splines for Medium-Term Investing 55(1) Highly Sensitive Splines for Short-Term Investing 56(2) Spline Regression Forecasting 58(3) Summary and Conclusions 61(2) Appendix: SAS® Program to Calculate Standard Error 63(2) Notes 65(2) References 67(2) About the Authors 69