Impacts of Dirty Data on Classification and Clustering Models

Since dirty data have negative influence on the accuracy of machine learning models, the relation between data quality and model results could be used in the selection of the proper model and data cleaning strategies. However, rare work has focused on this topic. Motivated by this, this chapter compares the impacts of missing, inconsistent, and conflicting data on basic classification and clustering models. Based on the evaluation observations, we suggest users how to select appropriate classification and clustering models and clean dirty data in the database and machine learning communities. Section 2.1 gives the research motivation of this chapter. Section 2.2 describes our assessment methodology. Section 2.3 presents our experimental results and analyses. We discuss the lessons learned from the evaluation in Sect. 2.4 and provide strategies of model selection and data cleansing, followed by a brief summary in Sect. 2.5.