A novel method based on long short term memory network and discrete-time zeroing neural algorithm for upper-limb continuous estimation using sEMG signals.
Yuanyuan ChaiKeping LiuChunxu LiZhongbo SunLong JinTian ShiPublished in: Biomed. Signal Process. Control. (2021)
Keyphrases
- estimation algorithm
- high accuracy
- computational cost
- cost function
- optimization algorithm
- detection algorithm
- improved algorithm
- dynamic programming
- experimental evaluation
- detection method
- preprocessing
- estimation accuracy
- segmentation method
- clustering method
- significant improvement
- k means
- computational complexity
- optimization method
- objective function
- support vector machine svm
- parameter estimation
- computationally efficient
- theoretical analysis
- recognition algorithm
- energy function
- matching algorithm
- convergence rate
- fitted q iteration
- correlation function
- probabilistic model
- tree structure
- input data
- junction tree
- synthetic and real images
- segmentation algorithm
- network architecture
- learning algorithm
- particle swarm optimization
- similarity measure
- continuous variables
- machine learning
- continuous wavelet transform
- shortest path algorithm
- estimation process
- optimal path
- image segmentation
- wireless sensor networks
- associative memory
- growing neural gas
- support vector machine
- classification algorithm
- markov chain
- expectation maximization