Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

Maxwell-Boltzmann Distribution: Problem Solving

Maxwell-Boltzmann Distribution: Problem Solving

Individual molecules in a gas move in random directions, but a gas containing numerous molecules has a predictable distribution of molecular speeds, which is known as the Maxwell-Boltzmann distribution, f(v).
This distribution function f(v) is defined by saying that the expected number N (v1,v2) of particles with speeds between v1 and v2 is given by

Atomic Nuclei: Nuclear Spin State Population Distribution

Atomic Nuclei: Nuclear Spin State Population Distribution

Near absolute zero temperatures, in the presence of a magnetic field, the majority of nuclei prefer the lower energy spin-up state to the higher energy spin-down state. As temperatures increase, the energy from thermal collisions distributes the spins more equally between the two states. The Boltzmann distribution equation gives the ratio of the number of spins predicted in the spin −½ (N−) and spin +½ (N+) states.

Mechanistic Models: Compartment Models in Algorithms for Numerical Problem Solving

Mechanistic Models: Compartment Models in Algorithms for Numerical Problem Solving

Mechanistic models play a crucial role in algorithms for numerical problem-solving, particularly in nonlinear mixed effects modeling (NMEM). These models aim to minimize specific objective functions by evaluating various parameter estimates, leading to the development of systematic algorithms. In some cases, linearization techniques approximate the model using linear equations.
In individual population analyses, different algorithms are employed, such as Cauchy's method, which uses a...

Entropy

Entropy

Salt particles that have dissolved in water never spontaneously come back together in solution to reform solid particles. Moreover, a gas that has expanded in a vacuum remains dispersed and never spontaneously reassembles. The unidirectional nature of these phenomena is the result of a thermodynamic state function called entropy (S). Entropy is the measure of the extent to which the energy is dispersed throughout a system, or in other words, it is proportional to the degree of disorder of a...

Reduced Mass Coordinates: Isolated Two-body Problem

Reduced Mass Coordinates: Isolated Two-body Problem

In classical mechanics, the two-body problem is one of the fundamental problems describing the motion of two interacting bodies under gravity or any other central force. When considering the motion of two bodies, one of the most important concepts is the reduced mass coordinates, a quantity that allows the two-body problem to be solved like a single-body problem. In these circumstances, it is assumed that a single body with reduced mass revolves around another body fixed in a position with an...

You might also read

Related Articles

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

Sort by

Same author

Free energy evaluation using marginalized annealed importance sampling.

Physical review. E·2022

Same author

Spatial Monte Carlo integration with annealed importance sampling.

Physical review. E·2021

Same author

A Generalization of Spatial Monte Carlo Integration.

Neural computation·2021

Same author

Statistical analysis of loopy belief propagation in random fields.

Physical review. E, Statistical, nonlinear, and soft matter physics·2015

Same author

Belief Propagation Algorithm for Portfolio Optimization Problems.

PloS one·2015

Same author

Susceptibility propagation by using diagonal consistency.

Physical review. E, Statistical, nonlinear, and soft matter physics·2013

Search research articles

Related Experiment Video

Updated: May 23, 2025

Closed-loop Neuro-robotic Experiments to Test Computational Properties of Neuronal Networks

Closed-loop Neuro-robotic Experiments to Test Computational Properties of Neuronal Networks

Published on: March 2, 2015

Dataset-free weight-initialization on restricted Boltzmann machine.

Muneki Yasuda¹, Ryosuke Maeno², Chako Takahashi¹

¹Graduate School of Science and Engineering, Yamagata University, Jonan 4-3-16, Yonezawa, 992-8510, Yamagata, Japan.

Neural Networks : the Official Journal of the International Neural Network Society

|March 7, 2025

Summary

Researchers developed a new dataset-free weight initialization method for Restricted Boltzmann Machines (RBMs). This method optimizes initial weight parameters using statistical mechanics to enhance learning efficiency in these probabilistic neural networks.

Keywords:

Learning parameter initialization Mean-field analysis Replica method Restricted Boltzmann machine

More Related Videos

Single-Molecule Tracking Microscopy - A Tool for Determining the Diffusive States of Cytosolic Molecules

Single-Molecule Tracking Microscopy - A Tool for Determining the Diffusive States of Cytosolic Molecules

Published on: September 5, 2019

Human Brown Adipose Tissue Depots Automatically Segmented by Positron Emission Tomography/Computed Tomography and Registered Magnetic Resonance Images

Human Brown Adipose Tissue Depots Automatically Segmented by Positron Emission Tomography/Computed Tomography and Registered Magnetic Resonance Images

Published on: February 18, 2015

Related Experiment Videos

Last Updated: May 23, 2025

Closed-loop Neuro-robotic Experiments to Test Computational Properties of Neuronal Networks

Closed-loop Neuro-robotic Experiments to Test Computational Properties of Neuronal Networks

Published on: March 2, 2015

Single-Molecule Tracking Microscopy - A Tool for Determining the Diffusive States of Cytosolic Molecules

Single-Molecule Tracking Microscopy - A Tool for Determining the Diffusive States of Cytosolic Molecules

Published on: September 5, 2019

Human Brown Adipose Tissue Depots Automatically Segmented by Positron Emission Tomography/Computed Tomography and Registered Magnetic Resonance Images

Human Brown Adipose Tissue Depots Automatically Segmented by Positron Emission Tomography/Computed Tomography and Registered Magnetic Resonance Images

Published on: February 18, 2015

Area of Science:

Artificial Intelligence
Machine Learning
Deep Learning

Background:

Dataset-free weight initialization methods like LeCun, Xavier, and He are established for feed-forward neural networks.
Such methods are currently underdeveloped for Restricted Boltzmann Machines (RBMs).

Purpose of the Study:

To derive a novel dataset-free weight-initialization method specifically for Bernoulli-Bernoulli RBMs.
To improve the learning efficiency of RBMs through optimized weight initialization.

Main Methods:

Statistical mechanical analysis was employed to derive the weight-initialization method.
Weight parameters are drawn from a Gaussian distribution with zero mean.
The standard deviation is optimized to maximize layer correlation (LC), derived via statistical mechanics.

Main Results:

A new dataset-free weight-initialization method for Bernoulli-Bernoulli RBMs was successfully derived.
The method involves drawing weights from a Gaussian distribution, with the standard deviation optimized for maximum layer correlation.
The proposed method aligns with Xavier initialization under specific conditions.

Conclusions:

The developed weight-initialization method offers an effective way to initialize RBMs without requiring training data.
Numerical experiments validated the method's effectiveness on both toy and real-world datasets.
This work addresses a gap in dataset-free initialization techniques for RBMs.