A New Hybrid Fault Prognosis Method for MFS Systems Based on Distributed Neural Networks and Recursive Bayesian Algorithm.
Mojtaba KordestaniMohammad Foad SamadiMehrdad SaifPublished in: IEEE Syst. J. (2020)
Keyphrases
- high accuracy
- cost function
- computational cost
- experimental evaluation
- significant improvement
- preprocessing
- clustering method
- objective function
- optimization algorithm
- detection method
- neural network
- synthetic and real images
- dynamic programming
- theoretical analysis
- input data
- single pass
- improved algorithm
- k means
- computationally efficient
- computational complexity
- tree structure
- distributed systems
- energy function
- classification method
- optimization method
- estimation algorithm
- segmentation algorithm
- recognition algorithm
- learning algorithm
- detection algorithm
- classification algorithm
- fuzzy art
- em algorithm
- expectation maximization
- segmentation method
- matching algorithm
- prior information
- similarity measure
- training process
- probabilistic model
- recursive algorithm
- search space
- machine learning methods
- neuro fuzzy systems
- convergence rate
- selection algorithm
- parameter estimation
- training algorithm
- support vector machine svm
- fault detection
- bit rate
- peer to peer
- feed forward neural networks
- camera network
- np hard
- artificial neural networks
- genetic algorithm