Results 61 to 70 of about 21,160 (284)
An extended class of multilayer perceptron [PDF]
Neurocomputing, 2008 In this work an extended class of multilayer perceptron is presented. This includes independent parameters, boundary conditions and lower and upper bounds. In some cases, such extensions contain a priori information of the problem. On some other situations they are necessary in order to define a correct representation for the solution. The use of
Roberto Lopez, Eugenio Oñate
openaire +2 more sources
Advanced Materials Technologies, EarlyView.This study shows that a lightweight blackbox neural network provides a practical, cost‐effective solution for bidirectional process prediction in laser‐induced graphene (LIG) fabrication. Achieving high predictive performance with minimal overhead, the approach democratizes machine learning (ML) for resource‐limited environments.
Maxim Polomoshnov +3 more
wiley +1 more source
Prunability of Multi-Layer Perceptrons Trained with the Forward-Forward Algorithm
MathematicsWe explore the sparsity and prunability of multi-layer perceptrons (MLPs) trained using the Forward-Forward (FF) algorithm, an alternative to backpropagation (BP) that replaces the backward pass with local, contrastive updates at each layer.
Mitko Nikov +2 more
doaj +1 more source
A Neural Network model with Bidirectional Whitening
, 2017 We present here a new model and algorithm which performs an efficient Natural gradient descent for Multilayer Perceptrons. Natural gradient descent was originally proposed from a point of view of information geometry, and it performs the steepest descent
H Park +3 more
core +1 more source
The Future of Research in Cognitive Robotics: Foundation Models or Developmental Cognitive Models?
Advanced Robotics Research, EarlyView.Research in cognitive robotics founded on principles of developmental psychology and enactive cognitive science would yield what we seek in autonomous robots: the ability to perceive its environment, learn from experience, anticipate the outcome of events, act to pursue goals, and adapt to changing circumstances without resorting to training with ...
David Vernon
wiley +1 more source
Design of real and complex recurrent neural networks for sound source localisation [PDF]
JASA Express LettersThis study explores a pruning technique based on the singular value decomposition for both real- and complex-valued recurrent neural networks (RNNs).
Vlad S. Paul, Philip A. Nelson
doaj +1 more source
Modeling Financial Time Series with Artificial Neural Networks [PDF]
, 2009 Financial time series convey the decisions and actions of a population of human actors over time. Econometric and regressive models have been developed in the past decades for analyzing these time series.
Versace, Massimiliano, Wong, Charles
core +1 more source
Reducing the Effects of Detrimental Instances
, 2014 Not all instances in a data set are equally beneficial for inducing a model of the data. Some instances (such as outliers or noise) can be detrimental. However, at least initially, the instances in a data set are generally considered equally in machine ...
Martinez, Tony, Smith, Michael R.
core +1 more source
Nonlocomotory Robotic Strategies for Dynamic Rotation Control in Terrestrial Robots: A Review
Advanced Robotics Research, EarlyView.Terrestrial robots increasingly require rapid body rotation to maintain stability and agility in complex environments. This review shows nonlocomotory rotational control strategies that operate without ground contact, including reaction wheels, tails, bars, limbs, and thrusters.
Y. Liang +14 more
wiley +1 more source
A Fast Learning Method for Multilayer Perceptrons in Automatic Speech Recognition Systems
Journal of Robotics, 2015 We propose a fast learning method for multilayer perceptrons (MLPs) on large vocabulary continuous speech recognition (LVCSR) tasks. A preadjusting strategy based on separation of training data and dynamic learning-rate with a cosine function is used to ...
Chenghao Cai +3 more
doaj +1 more source