Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Incremental hierarchical discriminant regression.

Juyang Weng1, Wey-Shiuan Hwang

  • 1Department of Computer Science and Engineering, Michigan State University, East Lansing, MI 48824, USA. weng@cse.msu.edu

IEEE Transactions on Neural Networks
|March 28, 2007
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Developmental Network-2: The Autonomous Generation of Optimal Internal-Representation Hierarchy.

IEEE transactions on neural networks and learning systems·2021
Same author

Learning to recognize while learning to speak: Self-supervision and developing a speaking motor.

Neural networks : the official journal of the International Neural Network Society·2021
Same author

Emergent neural turing machine and its visual navigation.

Neural networks : the official journal of the International Neural Network Society·2018
Same author

Editorial: Representation in the Brain.

Frontiers in psychology·2018
Same author

Motivated Optimal Developmental Learning for Sequential Tasks Without Using Rigid Time-Discounts.

IEEE transactions on neural networks and learning systems·2018
Same author

Modulation for emergent networks: serotonin and dopamine.

Neural networks : the official journal of the International Neural Network Society·2013
Same journal

Universal perceptron and DNA-like learning algorithm for binary neural networks: LSBF and PBF implementations.

IEEE transactions on neural networks·2013
Same journal

Guest editorial: special section on white box nonlinear prediction models.

IEEE transactions on neural networks·2011
Same journal

Data-based fault-tolerant control of high-speed trains with traction/braking notch nonlinearities and actuator failures.

IEEE transactions on neural networks·2011
Same journal

Guest editorial: special section on data-based control, modeling, and optimization.

IEEE transactions on neural networks·2011
Same journal

Neural network-based multiple robot simultaneous localization and mapping.

IEEE transactions on neural networks·2011
Same journal

Data-driven model-free adaptive control for a class of MIMO nonlinear discrete-time systems.

IEEE transactions on neural networks·2011
See all related articles

This study introduces incremental hierarchical discriminant regression (IHDR), a novel online learning system for high-dimensional data. IHDR efficiently models associative cortex development, handling dynamic data and memory challenges in real-time regression and decision tasks.

Area of Science:

  • Computational neuroscience
  • Machine learning
  • High-dimensional data analysis

Background:

  • Developing computational models for associative cortex development is crucial for understanding brain function.
  • Existing machine learning algorithms struggle with very high-dimensional spaces and dynamic data streams.
  • Online, real-time learning systems are needed for adaptive processing of complex sensory and motor information.

Purpose of the Study:

  • To present incremental hierarchical discriminant regression (IHDR), an online, real-time learning system for high-dimensional regression and decision spaces.
  • To provide a biologically motivated computational model for the automatic development of associative cortex.
  • To address challenges in parameter selection, memory loss, and varying sample sizes inherent in incremental tree construction.

Related Experiment Videos

Main Methods:

  • IHDR incrementally builds decision or regression trees using an online, real-time learning system.
  • It utilizes the output space to derive local subspaces spanned by discriminating features at each internal node.
  • A sample-size-dependent negative-log-likelihood (SDNLL) metric is employed to manage varying sample sizes across nodes.

Main Results:

  • The IHDR tree dynamically fits data with unknown distribution shapes and assigns long-term memory, mitigating memory loss.
  • It effectively handles large, small, and unbalanced sample-size cases through the SDNLL metric.
  • Experiments on synthetic data, face images, robot navigation video streams, and human-defined feature datasets demonstrate IHDR's capabilities.

Conclusions:

  • IHDR offers a robust and adaptive approach for real-time learning in very high-dimensional spaces.
  • Its biologically inspired design provides a computational model for associative cortex development.
  • The method effectively addresses key challenges in incremental tree-based learning, showing promise for diverse applications.