A Comprehensive Analysis of the Impact of Selecting the Training Set Elements on the Correctness of Classification for Highly Variable Ecological Data.
Adam KiersztynRafal LopuckiKrystyna KiersztynPawel KarczmarekPawel PowroznikDariusz CzerwinskiWitold PedryczPublished in: FUZZ-IEEE (2021)
Keyphrases
- training set
- training samples
- training data
- data sets
- data analysis
- data collection
- raw data
- classification algorithm
- synthetic data
- classification accuracy
- high quality
- database
- supervised learning
- class labels
- feature space
- feature vectors
- xml documents
- decision trees
- test data
- data distribution
- machine learning
- pattern classification
- classification rules
- training dataset
- functional magnetic resonance imaging
- text classification
- knowledge discovery
- data points
- prior knowledge
- support vector
- feature selection
- neural network