Motif discovery by the use of the complexity invariance

• Grant number: 13/16164-2
• Support Opportunities: Scholarships in Brazil - Master
• Start date: November 01, 2013
• End date: July 03, 2015
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Gustavo Enrique de Almeida Prado Alves Batista
• Grantee: Lucas Schmidt Cavalcante
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil

Abstract

Countless daily phenomena can be represented by time series, which are samples of a signal sorted by the time they were obtained. Motifs are subsequences, or patterns, of a time series that are recurrent. As a time series may have many motifs, it is necessary to attribute an importance to each one of them. In the literature there are two common definitions of the most important motif, often called the 1-motif. The first one defines that the motif is the most similar pair of subsequences; and the second one defines that it is the most recurrent subsequence. Recently it was proposed a new distance measure for time series, called Complexity-invariant Distance that takes into account the complexity of the series to measure their similarity. The goal of this work is to explore this new distance measure, as well as the concept of complexity for time series, to identify motifs of interest. The hypothesis of this work it is that 1-motifs with simple shapes such as ascending or descending line segments are hardly interesting motifs, no matter how similar they are or how many repetition they appear. In contrast, complex motifs usually represent subsequences of interest. This work explores the concept of complexity on motif discovery and ranking.

Academic Publications

(References retrieved automatically from State of São Paulo Research Institutions)

CAVALCANTE, Lucas Schmidt. Shapelets sampling and quality measurements. 2016. Master's Dissertation - Universidade de São Paulo (USP). Instituto de Ciências Matemáticas e de Computação (ICMC/SB) São Carlos.