Time-Series Filtering for Replicated Observations via a Kernel Approximate Bayesian Computation.
Takanori HasegawaKaname KojimaYosuke KawaiMasao NagasakiPublished in: IEEE Trans. Signal Process. (2018)
Keyphrases
- gaussian processes
- numerical integration
- kernel methods
- adaptive filtering
- bayesian networks
- efficient computation
- exact computation
- support vector
- efficiently computing
- non stationary
- fault tolerant
- bayesian learning
- information filtering
- kernel function
- maximum likelihood
- bayesian estimation
- pre computed
- dynamic time warping
- decision theory
- data mining tasks
- neural network
- probability distribution
- hidden markov models
- feature space
- feature selection