Improving expectation maximization algorithm over stellar data.
Hasan KurbanCan KockanMark JenneMehmet M. DalkilicPublished in: IEEE BigData (2017)
Keyphrases
- expectation maximization
- input data
- data sets
- noisy data
- em algorithm
- synthetic datasets
- synthetic data
- computational complexity
- matching algorithm
- parameter estimation
- learning algorithm
- data analysis
- detection algorithm
- prior information
- original data
- data collection
- incomplete data
- data reduction
- np hard
- prior knowledge
- spectral clustering
- bayesian framework
- information loss
- maximum likelihood estimation
- dimensional data
- maximum likelihood
- data distribution
- database
- neural network
- simulated annealing
- image data
- constrained optimization
- probabilistic model
- data sources
- k means
- association rules
- objective function