Strong optimality of the normalized ML models as universal codes and information in data.
Jorma RissanenPublished in: IEEE Trans. Inf. Theory (2001)
Keyphrases
- raw data
- historical data
- prior knowledge
- huge amounts
- data sets
- complex data
- information sources
- experimental data
- end users
- data mining tools
- sensor data
- training data
- log data
- collected data
- web data
- missing information
- computer systems
- information loss
- data sources
- data collection
- data processing
- sensitive information
- image data
- data analysis
- multiple sources
- information services
- database
- information resources
- domain knowledge
- prior information
- structural information
- statistical methods
- intrinsic characteristics
- statistical analysis
- model selection
- maximum likelihood
- accurate models
- stored data
- data repositories
- dimensional vector
- digital data
- privacy concerns
- private information
- missing values
- domain experts
- high dimensional data
- data mining techniques
- semi supervised
- information extraction
- probability distribution
- probabilistic model
- optimal solution