Missing Data in Distributed Analysis Sample Clauses

Missing Data in Distributed Analysis. The above-mentioned techniques of distributed analysis assume that the data are com- plete. However, it is especially common that data from multiple sources are subject to missing values. Based on our knowledge, ▇▇▇▇▇▇▇▇▇▇▇ and ▇▇▇▇▇▇ (2008) is the first and only paper that investigated missing data in a distributed analysis. In that paper, the author propose a privacy-preserving single imputation algorithm based on decision trees. The method can deal with the missing data problem when the data are collected from two sources and observed to have a univariate missing-data pattern.