NEURAL DATA ENVELOPMENT ANALYSIS: A SIMULATION
Luiz Biondi Neto, Marcos Teotonio Pereira, Estellita Lins, Eliane Gonçalves Gomes, João Carlos, Correia Baptista Soares de Mello, Fabiano S. Oliveira, Rua São Francisco Xavier, Rua Passo da Pátria · 2004
This paper investigates the creation of efficiency measurements structures of decision-making units (DMUs) called Neuro-DEA, by using high-speed optimization modules called Neuro-LP, inspired in the "philosophy " of an unconventional artificial neural network (ANN) and numerical methods. In addition, the linear programming problem LPP is transformed into an optimization problem without constraints by using a pseudo-cost function, where's added a term of penalty, causing high cost all time that one of the constraints goes violated. The problem is converted into a differential equations system. An unconventional ANN implements a numerical solution based gradient method. Significance: The convergence speed can be extremely high, if the proposed modules was integrated in a chip (VLSI – Very Large Scale Integration or CMOS- Complementary Metal-Oxide Semiconductor) and connected to a free slot in a computer.