Adding Seemingly Uninformative Labels Helps in Low Data Regimes.
Christos MatsoukasAlbert Bou I HernandezYue LiuKarin DembrowerGisele MirandaEmir KonukJohan Fredin HaslumAthanasios ZouzosPeter LindholmFredrik StrandKevin SmithPublished in: CoRR (2020)
Keyphrases
- data sets
- database
- data analysis
- data collection
- training data
- raw data
- computer systems
- synthetic data
- experimental data
- data structure
- complex data
- missing values
- high quality
- sensor data
- high dimensional data
- highly correlated
- low quality
- label information
- training samples
- data processing
- supervised learning
- small number
- probability distribution
- data sources
- prior knowledge
- relational databases
- clustering algorithm
- social networks
- neural network