Handling Missing Values when Applying Classification Models

By Saar-TsechanskyMaytal, ProvostFoster. 2007

80 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Jordan L. Boyd-Graber
2. He He
3. Brianna Satinoff
4. Alex Pentland
5. Bernardo Garcia-Bulle
6. Alejandro Noriega-Campero
7. Timothy C. Au
8. Michiel A. Bakker

Citations in the corpus

• Active Fairness in Algorithmic Decision Making (2018, methodology): “Similar methods can be derived for adapting logistic regressions to admit arbitrarily incomplete feature vectors [23, 26].”
• Besting the Quiz Master: Crowdsourcing Incremental Classification Games (2012, background): “2 Incremental classification is not equivalent to missing fea-2 tures, which have been studied at training time (Cesa-Bianchi et al., 2011), test time (Saar-Tsechansky and Provost, 2007), and in an online setting (Ros-tamizadeh et al., 2011).”
• Random Forests, Decision Trees, and Categorical Predictors: The "Absent Levels" Problem (2017, methodology): “These pseudo-instances are then sent down their appropriate daughter nodes, with the final prediction being derived from the weighted results of all the terminal nodes that are reached (Saar-Tsechansky and Provost, 2007).”
• Predictive Analytics in Information Systems Research (2011, unknown): “Missing values: Determining how to treat missing values depends on (1) whether the “missingness” is informative of the response (Ding and Simonoff 2010) and (2) whether the missing values are in the training set or in the to-be-predicted observations (Saar-Tsechansky and Provost 2007).”