New features in Version 2.2:

1.New Loss operators: Lmax, Logistic, Ltranche, Difference of two Losses, Buffered Probability of Exceedance.
2.New PSG Solvers: VANGRB,CARGRB,HELI. Note. These new solvers are available only if Gurobi Optimizer is installed on the user’s computer (see GUROBI).
3.New operations with matrices: Cutout Operation, Takein Operation
4.PSG MATLAB Environment:
•New PSG MATLAB Toolbox.
•New functions to create and solve problems in PSG Toolbox format: tbpsg_toolbox, tbpsg_vars, tbpsg_run, tbpsg_solution, tbpsg_matrix_pack, tbpsg_pmatrix_pack, tbpsg_vector_pack, tbpsg_point_pack, tbpsg_parameter_pack

 

New features in Version 2.0:

1.Short interface for problem statement is available for creation of optimization problems in Run-File and MATLAB Environments.
2.New PSG functions are included in version 2.0: Recourse Function, Spline Sum, function group of Kolmogorov-Smirnov distance, Koenker and Basset Error, Rockafellar Error, CVaR for Discrete Distribution as Function of Atom Probabilities, CVaR for Mixture of Normal Distributions as Function of Mixture Weights.
3.Possibility to construct cycles and implementation of cross-validation technique.
4.Function and constraint values (at some point) can be calculated with the new problem statement: Calculate.
5.PSG MATLAB Environment:
•Capabilities of subroutines Riskprog and Riskconstrprog are expanded.
•New functions: mpsg_setpathtoPSG, matrix_pack, pmatrix_pack, vector_pack, point_pack.

 

New features in Version 1.3:

1.Freeware PSG Express Edition 1.3 is available free of charge (can optimize up to 10 decision variables).
2.Types of variables can be specified (real, boolean, and integer) in all PSG environments. You can optimize various functions (e.g., Value-at-Risk) with variables of different type in an optimization problem.
3.External Functions are defined in PSG MATLAB Environment. You can program new functions and their gradient and include them to an optimization problem statement:
•External functions are implemented as mex-functions or m-functions (External Functions Interface).
•External functions (and their linear combinations) can be included in the objective function or in the constraints with mpsg_solver function.
•External functions are defined in riskprog, riskconstrprog functions.

4.  New Case Studies are added to the PSG version 1.3:

• Optimizing Intensity-Modulated Radiation Therapy Treatment Planning Problem (pm2_pen);
•Project Selection (fxchg, linear).

5. PSG MATLAB Environments has been modified in the following directions:

•New subroutines have been included in MATLAB version 1.3:
Riskprog, Riskconstrprog, Riskratioprog, riskparam, riskconstrparam, functionvalue, functionincrement, and functionsensitivity subroutines.
The first five subroutines are very similar to well known MATLAB functions linprog and quadrprog.
•New functions have been included in MATLAB version 1.3: mpsg_function_value, mpsg_function_increment, mpsg_function_sensitivity, mpsg_loss_values, mpsg_problem_exporttotext, mpsg_problem_exporttomat, and mpsg_problem_importfromtext.

 

New features in Version 1.2:

1.PSG MATLAB Environments has been modified:
•New subroutines have been included in MATLAB  version 1.2: Riskprog, Riskconstrprog, Riskratioprog, riskparam, riskconstrparam, functionvalue, functionincrement, and functionsensitivity subroutines. The first five subroutines are very similar to well known MATLAB functions linprog and quadrprog.
•New functions have been included in MATLAB  version 1.2: mpsg_function_value, mpsg_function_increment, mpsg_function_sensitivity, mpsg_loss_values, mpsg_problem_exporttotext, mpsg_problem_exporttomat, and mpsg_problem_importfromtext.
2.Thirteen new functions have been included in version 1.2: Polynomial Absolute, CVaR Component Positive, CVaR Component Negative, VaR Component Positive, VaR Component Negative, Maximum Component Positive, Maximum Component Negative, Quadratic function, Logarithms Sum, Logarithms Exponents Sum, Calculation of Partial Moment Two Penalty for Loss, Calculation of Partial Moment Two Penalty for Gain, Calculation of Partial Moment Two Deviation for Loss, Calculation of Partial Moment Two Deviation for Gain, Relative Entropy, Exponential Utility, Logarithmic Utility, Power Utility, Cardinality Positive, Cardinality Negative, Cardinality, Buyin Positive, Buyin Negative, Buyin, Fixed Charge Positive, Fixed Charge Negative, Fixed Charge.

 

 

 

 

 

 

 

 

 

© 2006-2016 American Optimal Decisions, Inc.