Probability Distribution Classification Processor

Abstract

This patent application discloses a system and method that preferably comprises several groups of artificial neural networks (ANNs) for classifying the probability distribution of random data. Each group of artificial neural networks is preferably trained to produce a selected output in response to data having a particular probability distribution, and each group preferably analyzes a different sample size of data. A parameter estimator module calculates statistical parameters for the different size data samples. The outputs of the several groups of artificial networks and of the parameter estimator are analyzed by a rule-based decision logic module which then selects the type of probability distribution that best describes the random data based on rules that correspond to ranges of values of the outputs of the artificial neural networks. (6 figures)

Open PDF

Document Details

Document Type
Technical Report
Publication Date
May 26, 2004
Accession Number
ADD020156

Entities

People

  • Christopher M. Deangelis

Organizations

  • United States Department of the Navy

Tags

Communities of Interest

  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Artificial Intelligence Software
  • Attorneys
  • Classification
  • Data Science
  • Data Sets
  • Distribution Functions
  • Estimators
  • Gaussian Distributions
  • Hidden Markov Models
  • Information Science
  • Machine Learning
  • Neural Networks
  • Probability
  • Probability Distributions
  • Recognition
  • Signal Processing
  • Statistics

Fields of Study

  • Computer science

Readers

  • Computational Modeling and Simulation
  • Computer Science/Computer Engineering/Data Science/Digital Signal Processing.
  • Neural Network Machine Learning.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Neural Networks