Wallenius Naïve Bayes

David Steinberg. One of the simplest methods for two-group classification is naïve Bayes, in which predictors are treated as though they provide independent information. Traditional event models underlying naive Bayes classifiers assume probability distributions that are not appropriate for binary data generated by human behavior. This paper develops a new event model, based on a … Continue reading Wallenius Naïve Bayes

Advertisements

FLAME: A Fast Large-scale Almost Matching Exactly Approach to Causal Inference

David Steinberg. This paper addresses a classical problem in causal inference: matching, where treatment units need to be matched to control units. Some of the main challenges in developing matching methods arise from the tension among (i) inclusion of as many covariates as possible in defining the matched groups, (ii) having matched groups with enough … Continue reading FLAME: A Fast Large-scale Almost Matching Exactly Approach to Causal Inference

Mind the Gap: Accounting for Measurement Error and Misclassification in Variables Generated via Data Mining

David Steinberg. Yang et al. consider the application of predictive data mining techniques in Information Systems research. Their focus is on the impact of data errors and misclassification on the subsequent data analysis by econometric models. Typically, data mining methods are first used to generate new variables (e.g., text sentiment), which are added into subsequent … Continue reading Mind the Gap: Accounting for Measurement Error and Misclassification in Variables Generated via Data Mining

The Surprising Power of Online Experiments

David Steinberg. One of the hot topics in internet commerce is A/B testing – the use of designed experiments to maximize revenue from web sites. The fact that experimental design is a great way to test ideas should not be a surprise to readers of this column. And many businesses have caught on to the … Continue reading The Surprising Power of Online Experiments