Abstract / Description of output
A practical method for calculating monotonic cubic smoothing splines is given. Linear sufficient conditions for monotonicity are employed, and the spline coefficients are obtained using quadratic programming. The method enables efficient cross-validation estimates of the smoothing parameter to be made and confidence
intervals to be calculated for the resulting spline. The results are easy to extend to histogram data.
intervals to be calculated for the resulting spline. The results are easy to extend to histogram data.
Original language | English |
---|---|
Pages (from-to) | 1126-1133 |
Journal | SIAM Journal on Scientific Computing |
Volume | 15 |
Issue number | 5 |
DOIs | |
Publication status | Published - 30 Sept 1994 |