scholarly journals Fast Threshold Optimization for Multi-Label Audio Tagging Using Surrogate Gradient Learning

2021 ◽  
Author(s):  
Thomas Pellegrini ◽  
Timothee Masquelier
Keyword(s):  
Threshold Optimization ◽  
Gradient Learning

A Q-learning based energy threshold optimization algorithm in LAA networks

2021 ◽  
pp. 1-1
Author(s):  
Errong Pei ◽  
Lineng Zhou ◽  
Bingguang Deng ◽  
Xun Lu ◽  
Zhizhong Zhang ◽  
...  
Keyword(s):  
Optimization Algorithm ◽  
Energy Threshold ◽  
Q Learning ◽  
Threshold Optimization

A weak derivative approach to optimization of threshold parameters in a multicomponent maintenance system

2001 ◽  
Vol 38 (02) ◽  
pp. 386-406 ◽  
Author(s):  
Bernd Heidergott
Keyword(s):  
Optimization Problems ◽  
Estimation Algorithm ◽  
Replacement Policy ◽  
Cost Performance ◽  
Weak Derivative ◽  
Preventive Replacement ◽  
Maintenance System ◽  
Threshold Optimization ◽  
Age Replacement ◽  
The Cost

We consider a multicomponent maintenance system controlled by an age replacement policy: when one of the components fails, it is immediately replaced; all components older than a threshold age θ are preventively replaced. Costs are associated with each maintenance action, such as replacement after failure or preventive replacement. We derive a weak derivative estimator for the derivative of the cost performance with respect to θ. The technique is quite general and can be applied to many other threshold optimization problems in maintenance. The estimator is easy to implement and considerably increases the efficiency of a Robbins-Monro type of stochastic approximation algorithm. The paper is self-contained in the sense that it includes a proof of the correctness of the weak derivative estimation algorithm.

A modified gradient learning algorithm with smoothing L1/2 regularization for Takagi–Sugeno fuzzy models

2014 ◽  
Vol 138 ◽  
pp. 229-237 ◽  
Author(s):  
Yan Liu ◽  
Wei Wu ◽  
Qinwei Fan ◽  
Dakun Yang ◽  
Jian Wang
Keyword(s):  
Learning Algorithm ◽  
Fuzzy Models ◽  
Gradient Learning ◽  
Takagi Sugeno

scholarly journals Training Restricted Boltzmann Machines With a D-Wave Quantum Annealer

2021 ◽  
Vol 9 ◽  
Author(s):  
Vivek Dixit ◽  
Raja Selvarajan ◽  
Muhammad A. Alam ◽  
Travis S. Humble ◽  
Sabre Kais
Keyword(s):  
Image Reconstruction ◽  
Graphical Model ◽  
Classification Performance ◽  
Supervised Machine Learning ◽  
Learning Performance ◽  
Restricted Boltzmann Machines ◽  
Log Likelihood ◽  
Improved Performance ◽  
Gradient Learning ◽  
D Wave

Restricted Boltzmann Machine (RBM) is an energy-based, undirected graphical model. It is commonly used for unsupervised and supervised machine learning. Typically, RBM is trained using contrastive divergence (CD). However, training with CD is slow and does not estimate the exact gradient of the log-likelihood cost function. In this work, the model expectation of gradient learning for RBM has been calculated using a quantum annealer (D-Wave 2000Q), where obtaining samples is faster than Markov chain Monte Carlo (MCMC) used in CD. Training and classification results of RBM trained using quantum annealing are compared with the CD-based method. The performance of the two approaches is compared with respect to the classification accuracies, image reconstruction, and log-likelihood results. The classification accuracy results indicate comparable performances of the two methods. Image reconstruction and log-likelihood results show improved performance of the CD-based method. It is shown that the samples obtained from quantum annealer can be used to train an RBM on a 64-bit “bars and stripes” dataset with classification performance similar to an RBM trained with CD. Though training based on CD showed improved learning performance, training using a quantum annealer could be useful as it eliminates computationally expensive MCMC steps of CD.

Sign in / Sign up

Export Citation Format

Share Document