Fall algorithm development using kinematic parameters measured from simulated falls performed in a quasi-realistic environment using accelerometry.
Alan K. BourkeMarc TorrentXavier ParraAndreu CatalàJohn NelsonPublished in: EMBC (2011)
Keyphrases
- preprocessing
- expectation maximization
- theoretical analysis
- learning algorithm
- high accuracy
- parameter selection
- recognition algorithm
- k means
- segmentation algorithm
- experimental evaluation
- input data
- convergence rate
- times faster
- clustering method
- detection algorithm
- parameter estimation
- worst case
- computational cost
- cost function
- optimal solution
- search space
- computational complexity
- neural network
- control parameters
- parameter settings
- image segmentation
- parameter space
- data structure
- matching algorithm
- optimization algorithm
- significant improvement
- dynamic programming
- em algorithm
- real time