Solving multiple objective programming problems using feed-forward artificial neural networks: the interactive FFANN procedure

Article Abstract:

A new interactive procedure for solving multiple objective programming problems is proposed. The technique, which uses feed-forward artificial neural networks (FFANN), is called the Interactive FFANN Procedure. The procedure calls for the decision maker to articulate preference information over representative samples from the nondominated set either by making pairwise comparisons similar to that in the Analytic Hierarchy Process or by assigning preference 'values' to the sample solutions. With this data, a FFANN can learn to represent the decision maker's preference structure. The FFANN also solves an optimization problem to search for improved solutions. The computational results show that the Interactive FFANN Procedure can make satisfactory solutions and is robust with regard to the neural network architecture.

Methods, Multiple criteria decision making, Computer networks, Neural networks

---

Using common random numbers for indifference-zone selection and multiple comparisons in simulation

Article Abstract:

Two experiment design and analysis procedures are presented that can simultaneously control the error in choosing the best among k systems and comparing this to other systems. These procedures minimize the sample size needed to achieve a fixed precision by taking advantage of the variance-reduction technique of common random numbers. One of the procedures, developed by Clark and Yang (1986), is statistically conservative since it is based on the Bonferroni inequality. The other is a new combined procedure that is precise under a specific dependence structure, but may sometimes be a little liberal. A numerical example is presented to demonstrate the application of the procedures.

Analysis, Simulation methods, Simulation, Numbers, Random, Random numbers

---

Similar abstracts:

• Abstracts: Scheduling School Buses. Modeling as constrained problem solving: an empirical study of the data modeling process
• Abstracts: A decomposition method for quadratic zero-one programming. Holt-Winters method with missing observations. Revising Forecats of Accounting Earnings: A Ccomparison with the Box-Jenkins Method
• Abstracts: Control: Organizational and Economic Approaches. Malaysian integrated population program performance: its relation to organizational and integration factors
• Abstracts: A Linear Dynamic Progrmming Approach to Irrigation System Management with Depleting Groundwater. Approximation Methods for the Uncapacitated Dynamic Lot Size Problem

This website is not affiliated with document authors or copyright owners. This page is provided for informational purposes only. Unintentional errors are possible.