Time-series data mining

Abstract

In almost every scientific field, measurements are performed over time. These observations lead to a collection of organized data called time series. The purpose of time series data mining is to try to extract all meaningful knowledge from the shape of data. Even if humans have a natural capacity to perform these tasks, it remains a complex problem for computers. In this paper we intend to provide a survey of the techniques applied for time series data mining. The first part is devoted to an overview of the tasks that have captured most of the interest of researchers. Considering that in most cases, time series task relies on the same components for implementation, we divide the literature depending on these common aspects, namely representation techniques, distance measures and indexing methods. The study of the relevant literature has been categorized for each individual aspects. Four types of robustness could then be formalized and any kind of distance could then be classified. Finally, the study submit various research trends and avenues that can be explored in the near future. We hope that this paper can provide a broad and deep understanding of the time series data mining research field.

REBASS: Recomendação de Objetos de Aprendiza- gem Utilizando Similaridade de Sessões

“…Segundo Esling e Agon [26], em quase todas as áreas da ciência há a necessidade de se realizar medições de variáveis ao longo de um perí-odo de tempo, o que leva a criação destes conjuntos de dados ordenados.…”

“…Considerando-se a definição de série temporal dada por Esling e Agon [26] e Fu [11], percebe-se que as sequências de OAs acessados por um usuário durante uma sessão de consulta a um repositório podem ser descritas como uma série temporal, desde que estejam cronologicamente ordenadas e os aspectos relevantes para a análise de similaridade entre as sequências puderem ser modelados com funções matemáticas. A Figura 1 ilustra esta correlação.…”

Developing a pattern discovery method in time series data and its GPU acceleration

“…Time series data is a series of values sampled at time intervals [1]. Time series data captures a lot of information including inter-patterns, trends, and correlations.…”

Robust histogram-based feature engineering of time series data

“…The operations that can be performed based on the time series data are mainly directed towards pattern discovery, clustering, classification and rule discovery [2]. Due to the density of the available data that can be collected by the sensors and the nature of the time series data, there are three main tasks that need to be defined so that the previously mentioned operations can be performed [3]: Dimensionality reduction and Data representation, Distance measurement and Indexing.…”

“…There are several well known distance measurements that can be used to identify difference between time series. The distance measurement must be invariant to many transformations of the time series data such as amplitude or time shifting, uniform amplification, additive noise, time scaling, etc [3]. For this reason many types of distance measurements have been proposed in the literature and each have their advantages and drawbacks.…”