Evaluating discrete choice prediction models when the evaluation data is corrupted: analytic results and bias corrections for the area under the ROC.
Roger M. SteinPublished in: Data Min. Knowl. Discov. (2016)
Keyphrases
- data sets
- data analysis
- prior knowledge
- predictive model
- high quality
- incomplete data
- data processing
- historical data
- data sources
- data distribution
- synthetic data
- statistical analysis
- discrete data
- prediction error
- learning models
- raw data
- neural network
- data collection
- training data
- knowledge discovery
- database
- data structure
- prediction model
- models built
- decision trees
- roc curve
- probability distribution
- data points
- missing values
- statistical models
- model selection
- prediction accuracy
- high dimensional data