Forecasting Method for Grouped Time Series with the Use of k-Means Algorithm.
Nadezhda N. AstakhovaLiliya A. DemidovaEvgeny V. NikulchevPublished in: CoRR (2015)
Keyphrases
- k means
- clustering method
- high accuracy
- cost function
- dynamic programming
- preprocessing
- improved algorithm
- detection algorithm
- computational cost
- significant improvement
- objective function
- segmentation algorithm
- optimization algorithm
- input data
- experimental evaluation
- detection method
- optimal solution
- expectation maximization
- recognition algorithm
- computationally efficient
- segmentation method
- optimization method
- estimation algorithm
- computational complexity
- data clustering
- em algorithm
- hierarchical clustering
- matching algorithm
- classification algorithm
- support vector machine svm
- probabilistic model
- learning algorithm
- tree structure
- energy function
- theoretical analysis
- similarity measure
- spectral clustering
- pairwise
- neural network
- initial cluster centers
- cluster analysis
- convergence rate
- clustering algorithm
- cluster centers
- fuzzy clustering algorithm
- grey prediction
- box jenkins