Does Size Matter - How Much Data is Required to Train a REG Algorithm?
Mariët TheuneRuud KoolenEmiel KrahmerSander WubbenPublished in: ACL (2) (2011)
Keyphrases
- experimental evaluation
- noisy data
- input data
- high quality
- cost function
- memory space
- data collection
- computational cost
- k means
- computational complexity
- data analysis
- dynamic programming
- optimal solution
- data reduction
- optimization algorithm
- image data
- particle swarm optimization
- original data
- information loss
- data distribution
- matching algorithm
- synthetic datasets
- single scan
- objective function
- probabilistic model
- significant improvement
- search algorithm
- convergence rate
- data sets
- scales linearly
- space requirements
- dimensional data
- space complexity
- incomplete data
- data quality
- prior information
- np hard
- tree structure
- probability distribution
- data mining techniques
- recognition algorithm
- data structure
- high dimensional data
- training data
- window size
- detection algorithm
- fixed size
- decision trees
- knowledge discovery
- worst case