In this paper we present a new framework based on the existing Laplacian SVM [14], by adding a regularization term to the standard optimization problem and solving the new optimization equation derived from there.
The solubility of the test protein is predicted to increase if s j >0 and decrease if s j < 0. The standard optimization model used by SVM does not impose any bounds on the weights w t.
Since the complexity of such constraints and the switching instants is unknown, it is difficult to solve the problem by standard optimization techniques.
Fitting these two (non-linear) parameters by a standard optimization technique like the Marquardt-Levenberg algorithm is very slow and hence not suitable for processing an acquisition of 9 images containing 496×658 pixels.
The proposed method can obtain solutions with smooth boundaries and the parameterization allows solution by standard optimization methods.
Since the optimization model requires computing the interpolation variances of blocks being simulated/estimated in each iteration, the problem cannot be solved by standard optimization tools.
This deviation can be numerically computed by employing standard optimization techniques such as SQP and HGA.
Therefore, the optimal Lagrange multipliers of μ ∗ and λ ∗, in which maximize d, can be obtained by using the standard convex optimization techniques [23].
As it is shown in the numerical investigation presented in the next section, the resulting optimization models are easily tractable by use of standard optimization packages.
Our approach is based on the use of mixed-integer dynamic optimization (MIDO) models that minimize the Akaike criterion, and that can be solved by standard optimization algorithms.
