A VLSI optimal constructive algorithm for classification problems

Conference · Wed Oct 01 00:00:00 EDT 1997

OSTI ID:532561

Beiu, V ^[1]; Draghici, S; Sethi, I K ^[2]

---

1. Los Alamos National Lab., NM (United States)
2. Wayne State Univ., Detroit, MI (United States)

If neural networks are to be used on a large scale, they have to be implemented in hardware. However, the cost of the hardware implementation is critically sensitive to factors like the precision used for the weights, the total number of bits of information and the maximum fan-in used in the network. This paper presents a version of the Constraint Based Decomposition training algorithm which is able to produce networks using limited precision integer weights and units with limited fan-in. The algorithm is tested on the 2-spiral problem and the results are compared with other existing algorithms.

• Research Organization: Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
• Sponsoring Organization: USDOE Assistant Secretary for Human Resources and Administration, Washington, DC (United States)
• DOE Contract Number: W-7405-ENG-36
• OSTI ID: 532561
• Report Number(s): LA-UR-97-1609; CONF-971072-1; ON: DE97008305; TRN: AHC29721%%111
• Resource Relation: Conference: 7. international conference on artificial neural networks, Lausamme (Switzerland), 7-10 Oct 1997; Other Information: PBD: [1997]
• Country of Publication: United States
• Language: English

Related Subjects

99 MATHEMATICS
COMPUTERS
INFORMATION SCIENCE
MANAGEMENT
LAW
MISCELLANEOUS
NEURAL NETWORKS
INTEGRATED CIRCUITS
ALGORITHMS
IMPLEMENTATION
COMPARATIVE EVALUATIONS
COMPUTER ARCHITECTURE
OPTIMIZATION