Partitioning Interval-Valued Data Using Regression

• Activity Number: 313 - Recent Advances in Symbolic Data Analysis
• Type: Topic-Contributed
• Date/Time: Wednesday, August 11, 2021 : 3:30 PM to 5:20 PM
• Sponsor: Section on Statistical Learning and Data Science
• Author(s): Lynne Billard* and Fei Liu
• Companies: University of Georgia and Bank of America
• Keywords: intervals; clustering

Abstract

While there has been a lot of work using regression-based algorithms to partition a data set into clusters for classical data, no such algorithms have been published for a data set of interval-valued observations. A new algorithm is proposed based on the k-means algorithm of MacQueen (1967) and the dynamical partitioning method of Diday (1973) and Diday and Simon (1976). The partitioning criterion is based on establishing regression models, with Hausdorff, city-block or center distances used as measures between the underlying regression models in each sub-cluster. Simulation studies show the proposed algorithm is superior to the traditional k-means method.