Backpropagation Neural Network as Earthquake Early Warning
Tool using a new Elementary Modified Levenberg–Marquardt
Algorithm to minimise Backpropagation Errors

Lin, Jyh-Woei; Chao, Chia‐Ter

doi:10.5194/gi-2018-13

Cited by 3 publications

(4 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The forward-propagation process and the back-propagation process of the BPNN algorithm constitute the common process of error propagation. In the forward-propagation process of the signal, it is necessary to provide input samples to each neuron of the input layer, calculate the net input and output of the output layer and each hidden layer, and then calculate the prediction results of the neural network [ 14 , 15 ]. If there is a large error between the calculated prediction result and the expected output, the back-propagation of the error is started.…”

Section: Automatic Tqa Modelmentioning

confidence: 99%

See 1 more Smart Citation

Unreferenced English articles’ translation quality-oriented automatic evaluation technology using sparse autoencoder under the background of deep learning

2022

PLoS ONE

View full text Add to dashboard Cite

Currently, both manual and automatic evaluation technology can evaluate the translation quality of unreferenced English articles, playing a particular role in detecting translation results. Still, their deficiency is the lack of a close or noticeable relationship between evaluation time and evaluation theory. Thereupon, to realize the automatic Translation Quality Assessment (TQA) of unreferenced English articles, this paper proposes an automatic TQA model based on Sparse AutoEncoder (SAE) under the background of Deep Learning (DL). Meanwhile, the DL-based information extraction method employs AutoEncoder (AE) in the bilingual words’ unsupervised learning stage to reconstruct the translation language vector features. Then, it imports the translation information of unreferenced English articles into Bilingual words and optimizes the extraction effect of language vector features. Meantime, the translation language vector feature is introduced into the automatic DL-based TQA. The experimental findings corroborate that when the number of sentences increases, the number of actual translation errors and the evaluation scores of the proposed model increase, but the Bilingual Evaluation Understudy (BLEU) score is not significantly affected. When the number of sentences increases from 1,000 to 6,000, the BLEU increases from 96 to 98, which shows that the proposed model has good performance. Finally, the proposed model can realize the high-precision TQA of unreferenced English articles.

show abstract

Section: Automatic Tqa Modelmentioning

confidence: 99%

“…In Eq (15), β-the weight coefficient to control the sparsity penalty. After the loss function J SAE (θ) is minimized, the parameter θ can be obtained.…”

Section: Plos Onementioning

confidence: 99%

Unreferenced English articles’ translation quality-oriented automatic evaluation technology using sparse autoencoder under the background of deep learning

2022

PLoS ONE

View full text Add to dashboard Cite

show abstract

“…It was multi-layer feed forward nets and used chain rule in iterative manner for calculating the gradient function for each layer. It important to call a suitable activation function for each layer [11][12][13][14].…”

Section: Introductionmentioning

confidence: 99%

Application and evaluation of the neural network in gearbox

Al-Azzawi

2020

TELKOMNIKA

View full text Add to dashboard Cite

We developed old designed of a Back-Propagation neural network (BPNN), which it was designed by other researchers, and we made modification in their structure. The 1 st velocity ratio was discriminated by lowest speed, and highest twist. The 6 th velocity ratio was discriminated by highest speed, and lowest twist. The aim of this paper is to design neural structure get best performance to control an electrical automotive transportation six-speed gearbox of the vehicle. We focus on the evaluation of the BPNN to select the suitable number of layers and neurons. Experimentally, the structure of the proposed BPNN are constructed from four layers: eight input nodes in the first layer that received data in binary number, 45 neurons in 1 st hidden-layer, 25 neurons in 2 nd hidden-layer, and 6 neurons in the fourth layer. The MSE and number of Epochs are the main factors used for the evaluation of the proposed structure, and compared with the other structures which was designed by other researchers. Experimentally, we discovered that the best value of Epoch and MSE was chosen when the BPNN consisted of two hidden-layers, 45, and 25 neurons in the 1 st and 2 nd hidden-layer respectively. The implementation was applied using MATLAB software.

show abstract

“…It is a generalization of the delta rule to multi-layered feed-forward networks, made possible by using the chain rule to iteratively compute gradients for each layer. Back-propagation requires an activation function ( that is used by the artificial neurons ) to be differentiable [10][11] [12].…”

Section: Introductionmentioning

confidence: 99%

Simulation design of an Intelligent system for Automotive transmission Gearbox Based on FPGA

Saeed¹

2018

emitter

View full text Add to dashboard Cite

In this paper, an artificialÂ intelligent system has been designed, realized, and downloaded intoÂ FPGA (Field Programmable Gate Array), which is used to control five speed ratio steps ( 1,2,3,4,5) of an electrically controlled type ofÂ automotive transmission gearbox of a vehicle, the first speed ratio step (1) is characterized by theÂ highest torque, a lowest velocity, while, theÂ fifth step is characterized by the lowest torque, and highest velocity.The Back-propagation neural network has been used as the intelligent system for the proposed system. The proposed neural network is composed fromÂ Â eight neurons in the input layer, five neurons in the hidden layer, and five neurons in the output layer. For real downloading into the FPGA, Satlins and Satlin linear activation function has been used for the hidden and output layers respectively. The training function Trainlm ( Levenberg-Marqurdt training) has been used as a learning method for the proposed neural network, which it has a powerful algorithm.Â The proposed simulation system has been designed and downloaded into the FPGA using MATLAB and ISE Design Suit software packages.

show abstract

Backpropagation Neural Network as Earthquake Early Warning Tool using a new Elementary Modified Levenberg–Marquardt Algorithm to minimise Backpropagation Errors

Cited by 3 publications

References 19 publications

Unreferenced English articles’ translation quality-oriented automatic evaluation technology using sparse autoencoder under the background of deep learning

Unreferenced English articles’ translation quality-oriented automatic evaluation technology using sparse autoencoder under the background of deep learning

Application and evaluation of the neural network in gearbox

Simulation design of an Intelligent system for Automotive transmission Gearbox Based on FPGA

Contact Info

Product

Resources

About