Time-delay neural networks: representation and induction of finite-state machines.

Time-delay neural networks: representation and induction of finite-state machines. Research Abstract Details 

Research Abstract Table of Contents

Jump to the:

Time-delay neural networks: representation and induction of finite-state machines. Abstract Text:

D S Clouse,C L Giles,B G Horne,G W Cottrell,

In this work, we characterize and contrast the capabilities of the general class of time-delay neural networks (TDNNs) with input delay neural networks (IDNNs), the subclass of TDNNs with delays limited to the inputs. Each class of networks is capable of representing the same set of languages, those embodied by the definite memory machines (DMMs), a subclass of finite-state machines. We demonstrate the close affinity between TDNNs and DMM languages by learning a very large DMM (2048 states) using only a few training examples. Even though both architectures are capable of representing the same class of languages, they have distinguishable learning biases. Intuition suggests that general TDNNs which include delays in hidden layers should perform well, compared to IDNNs, on problems in which the output can be expressed as a function on narrow input windows which repeat in time. On the other hand, these general TDNNs should perform poorly when the input windows are wide, or there is little repetition. We confirm these hypotheses via a set of simulations and statistical analysis.

Time-delay neural networks: representation and induction of finite-state machines. Publishing Authors By Initials

DS Clouse,CL Giles,BG Horne,GW Cottrell,

For similar abstracts research abstracts see: abstracts research

PUBMED ID PMID:

MEDLINE DATE:

Time-delay neural networks: representation and induction of finite-state machines. Journal Published:

PUBLICATION TYPE: Journal Article

Journal: IEEE transactions on neural networks / a publicati

VOLUME: 8

Page Numbers: 1065-70

Journal Abbreviation:

ISSN: 1045-9227

DAY: 7

MONTH: 02

YEAR: 1997

Time-delay neural networks: representation and induction of finite-state machines. Information

Number of References:

LANGUAGE: eng

NlmUniqueID: 101211035

Time-delay neural networks: representation and induction of finite-state machines. Keywords Mesh Terms:

KEYWORDS:

MESH TERMS:

Chemical & Substance for Abstract: Time-delay neural networks: representation and induction of finite-state machines. Information

Substance Name:

Registry Number:

Grant and Affiliation Information for Time-delay neural networks: representation and induction of finite-state machines.

AFFILIATION: California Univ., San Diego, La Jolla, CA.

Country: United States

AGENCY:

GRANT:

ACRONYM:

MEDLINETA: IEEE Trans Neural Netw

REFSOURCE:

DATABASENAME:

ACCESSION NUMBER:

Number Hits: 0

Time-delay neural networks: representation and induction of finite-state machines Related Publications

• ACETYLCHOLINE AND 5-HYDROXYTRYPTAMINE IN THE SNAIL BRAIN.
• Adverse effects of toxins and drugs on the surfactant systems.
• Alveolar distraction osteogenesis for implant site development.
• Analysis of the fractional hilbert transform.
• BMI-based body size guides for women and men: development and validation of a novel pictorial method to assess weight-related concepts.
• Bessel function output from an optical correlator.
• Candida glabrata endophthalmitis following penetrating keratoplasty.
• Cerebral oxygenation.
• Compilation of United States nuclear standards 4th edition 1967. ORNL-NSIC-43.
• Compilation of United States nuclear standards. 5th edition 1968. ORNL-NSIC-57.
• Compilation of national and international nuclear standards (excluding U.S. activities). ORNL-NSIC-38.
• Compilation of national and international nuclear standards (excluding U.S. activities). ORNL-NSIC-50.
• Effect of alcohols on the stability of iprodione in solution.
• Green laser light (532nm) activates a chloride current in the C1 neuron of Helix aspersa.
• Hepatocyte growth factor-regulated tyrosine kinase substrate (HRS) mediates post-endocytic trafficking of protease-activated receptor 2 and calcitonin receptor-like receptor.
• High altitude and retinal detachment.
• High diffraction efficiency from one- and two-dimensional Nyquist frequency binary-phase gratings.
• Newborns and anesthetic neurotoxicity.
• Notched rotation osteotomy; a method employed in the correction of torsion of the tibia and other conditions.
• Pregnancy in thyrotoxicosis under treatment with thiouracil.
• Role of Social Support, Hardiness, and Acculturation as Predictors of Mental Health among International Students of Asian Indian Origin.
• Septate vagina complicated by pregnancy.
• Specificity of cardiolipin antigens in tests for syphilis.
• Teaching for co-ordinated curative and preventive medicine in Europe today.
• The roles of visual expertise and visual input in the face inversion effect: Behavioral and neurocomputational evidence.
• Tilting at windmills.
• Time-delay neural networks: representation and induction of finite-state machines.
• Vicarious trauma:predictors of clinicians' disrupted cognitions about self-esteem and self-intimacy.
• Why is the fusiform face area recruited for novel categories of expertise? A neurocomputational investigation.
• [Not Available.]