Introduction to Neural Network and Improved Algorithm to Avoid Local Minima and Faster Convergence

Bari, Ayesha; Bhasin, Kavleen; Karnawat, Darshan N.

doi:10.1007/978-3-642-25734-6_61

Cited by 2 publications

(3 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…After the roller coaster running a circulation, load the data from SD card and send to the upper computer. According to the real running condition of the roller coaster, the system employs an evolutional wavelet based denoising algorithms [7]. Six kinds of characteristic values are extracted: the peak indicators, the valid values, the magnitude margin, the kurtosis values, the pulse indicators and the center of gravity of spectrum.…”

Section: Program In Dspmentioning

confidence: 99%

See 1 more Smart Citation

Fault Diagnosis System for Roller Coaster Based on Acceleration Signals

Zhu²

2012

AMM

View full text Add to dashboard Cite

Based on the requirements of special equipment safety and the current status of recreation facilities test system, a fault diagnosis system for roller coaster was designed. The signal acquisition device was designed, taking digital signal processor as the main control chip. High-frequency vibration acceleration signal and low-frequency motion acceleration signal were acquired. And the data were transmitted to upper computer. By using LabVIEW, the software of upper computer was developed. FFT analysis and wavelet analysis were applied to process the data. BP neural network was applied to fault diagnosis. The test result shows that the system is quick, accurate and reliable.

show abstract

Section: Program In Dspmentioning

confidence: 99%

“…The fault diagnosis is completed by the BP neural network [9]. Firstly, the neural network should be trained.…”

Section: Simulationmentioning

confidence: 99%

Fault Diagnosis System for Roller Coaster Based on Acceleration Signals

Zhu²

2012

AMM

View full text Add to dashboard Cite

show abstract

“…Overly large networks tend to overfit problems at the expense of generalization, while networks with an insufficient numbers of neurons or synapses are unable to encode sufficiently complex mappings, leading to suboptimal performance [21]. Evaluating architectures can also be problematic as the typical gradient descent algorithms used to train the networks are inconsistent as they can become trapped in local minima of the error space [2].…”

Section: Introductionmentioning

confidence: 99%

Improving the Performance of CGPANN for Breast Cancer Diagnosis Using Crossover and Radial Basis Functions

Manning

Walsh

2013

Evolutionary Computation, Machine Learning and Data Mining in Bioinformatics

View full text Add to dashboard Cite

Abstract.Recently published evaluations of the topology and weight evolving artificial neural network algorithm Cartesian genetic programming evolved artificial neural networks (CGPANN) have suggested it as a potentially powerful tool for bioinformatics problems. In this paper we provide an overview of the CGPANN algorithm and a brief case study of its application to the Wisconsin breast cancer diagnosis problem. Following from this, we introduce and evaluate the use of RBF kernels and crossover to CGPANN as a means of increasing performance and consistency.

show abstract

Introduction to Neural Network and Improved Algorithm to Avoid Local Minima and Faster Convergence

Cited by 2 publications

References 2 publications

Fault Diagnosis System for Roller Coaster Based on Acceleration Signals

Fault Diagnosis System for Roller Coaster Based on Acceleration Signals

Improving the Performance of CGPANN for Breast Cancer Diagnosis Using Crossover and Radial Basis Functions

Contact Info

Product

Resources

About