The performance of support vector regression estimation was studied. It is found that the insensitive factor ε

penalty factor

and the kernel function along with its parameter are the main factors affecting the performance of support vector regression estimation.It remains a critical unsolved problem to determine the parmaeters of SVM. Cross-validation methods are commonly used in practice to decide the parameters of SVM

but they are usually expensive in computing time. A novel adaptive support vector machine (A-SVM) was proposed to determine the optimal parameters adaptively. The algorithms for adaptively tuning parameters of SVM were worked out. A-SVM was successfully applied in modeling delayed coking process. Compared with RBFN-PLSR methods

A-SVM was superior in both fitting accuracy and prediction performance. The proposed algorithms in general may be used in modeling complex chemical processes.