Time series analysis

Time series analysis are considered as a key component of strategic control over a wide range of disciplines in business, science and engineering. Time series data is a sequence of observations collected over intervals of time. Each time series describes a phenomenon as a function of time. Analysis of time series data includes detecting trends (or patterns) in a time series sequence. In recent years, data mining appeared and was recognized as a new technology for data analysis. Data mining is the process of discovering potentially valuable patterns, associations, movements, sequences and dependencies in the data. Extraction techniques may reveal information that many traditional business analysis and statistical techniques fail to deliver. In this paper, we will adapt and innovate data mining techniques for analysis of time series data. By using data mining techniques, the maximum frequent patterns are discovered and used in predicting future trends or sequences where trends describe the behavior of a sequence. In order to include different types of time series (eg, irregular and non-systematic), we consider models of the past often the same time sequences (local transactions) and other "dependent" time sequences (global models). We use the word "dependent" rather than the word "similar" to an emphasis on real life time series, where two time series sequences can be quite different (in values, forms, ... etc.) but they still react the same way depending on conditions.

In this paper, we propose Dependency Mining technique that can be used in predicting time series sequences. The proposed technique consists of three phases: (a) for all time series sequences, their trend generated sequences, (b) often reveal trends models generate pattern vectors (to keep the trend of frequent information patterns), use the template vectors trend to predict future series sequences.