Initialization Using Perlin Noise for Training Networks with a Limited Amount of Data.
Nakamasa InoueEisuke YamagataHirokatsu KataokaPublished in: ICPR (2020)
Keyphrases
- raw data
- data collection
- high quality
- statistical analysis
- data processing
- data mining techniques
- noisy data
- knowledge discovery
- synthetic data
- network structure
- training samples
- missing data
- labelled data
- original data
- data quality
- network resources
- database
- prior knowledge
- data structure
- training data
- databases
- data sets
- data sources
- hidden markov models
- training examples
- training set
- attribute values
- data analysis
- statistical methods
- social networks
- data mining
- training dataset
- random noise