Abstract. A probabilistic similarity query over uncertain data assigns to each uncertain database object o a probability indicating the likelihood that o meets the query predicate. In this paper, we formalize the notion of uncertain time series and introduce two novel and important types of probabilistic range queries over uncertain time series. Furthermore, we propose an original approximate representation of uncertain time series that can be used to efficiently support both new query types by upper and lower bounding the Euclidean distance.

Abstract. Similarity search in time series data is required in many application fields. The most prominent work has focused on similarity search considering either complete time series or similarity according to subsequences of time series. For many domains like financial analysis, medicine, environmental meteorology, or environmental observation, the detection of temporal dependencies between different time series is very important. In contrast to traditional approaches which consider the course of the time series for the purpose of matching, coarse trend information about the time series could be sufficient to solve the above mentioned problem. In particular, temporal dependencies in time series can be detected by determining the points of time at which the time series exceeds a specific threshold. In this paper, we introduce the novel concept of threshold queries in time series databases which report those time series exceeding a user-defined query threshold at similar time frames compared to the query time series. We present a new efficient access method which uses the fact that only partial information of the time series is required at query time. The performance of our solution is demonstrated by an extensive experimental evaluation on real world and artificial time series data.

Similarity search in time series data is an active area of research. In this paper, we introduce the novel concept of threshold-similarity queries in time series databases which report those time series exceeding a user-defined query threshold at similar time frames compared to the query time series. In addition, we present a new data structure to support threshold similarity queries efficiently. The performance of our solution is demonstrated by an extensive experimental evaluation.

The analysis of time series data is of capital importance for pharmacogenomics since the experimental evaluations are usually based on observations of time dependent reactions or behaviors of organisms. Thus, data mining in time series databases is an important instrument towards understanding the effects of drugs on individuals. However, the complex nature of time series poses a big challenge for effective and efficient data mining. In this paper, we focus on the detection of temporal dependencies between different time series: we introduce the novel analysis concept of threshold queries and its semi-supervised extension which supports the parameter setting by applying training datasets. Basically, threshold queries report those time series exceeding an user-defined query threshold at certain time frames. For semi-supervised threshold queries the corresponding threshold is automatically adjusted to the characteristics of the data set, the training dataset, respectively. In order to support threshold queries efficiently, we present a new efficient access method which uses the fact that only partial information of the time series is required at query time. In an extensive experimental evaluation we demonstrate the performance of our solution and show that semi-supervised threshold queries applied to gene expression data are very worthwhile.

Abstract. Effective similarity search in multi-media time series such as video or audio sequences is important for content-based multi-media retrieval applications. We propose a framework that extracts a sequence of local features from large multi-media time series that reflect the characteristics of the complex structured time series more accurately than global features. In addition, we propose a set of suitable local features that can be derived by our framework. These features are scanned from a time series amplitude-levelwise and are called amplitude-level features. Our experimental evaluation shows that our method models the intuitive similarity of multi-media time series better than existing techniques.

Abstract. Similarity search in time series data is used in diverse domains. The most prominent work has focused on similarity search considering either complete time series or certain subsequences of time series. Often, time series like temperature measurements consist of periodic patterns, i.e. patterns that repeatedly occur in defined periods over time. For example, the behavior of the temperature within one day is commonly correlated to that of the next day. Analysis of changes within the patterns and over consecutive patterns could be very valuable for many application domains, in particular finance, medicine, meteorology and ecology. In this paper, we present a framework that provides similarity search in time series databases regarding specific periodic patterns. In particular, an efficient threshold-based similarity search method is applied that is invariant against small distortions in time. Experiments on real-world data show that our novel similarity measure is more meaningful than established measures for many applications.

In the past decade, many automated prediction methods for the subcellular localization of proteins have been proposed, utilizing a wide range of principles and learning approaches. Based on an experimental evaluation of different methods and on their theoretical properties, we propose to combine a well balanced set of existing approaches to new, ensemble-based prediction methods. The experimental evaluation shows our ensembles to improve substantially over the underlying base methods.

Abstract. Effective and efficient data mining in time series databases is essential in many application domains as for instance in financial analysis, medicine, meteorology, and environmental observation. In particular, temporal dependencies between time series are of capital importance for these applications. In this paper, we present TQuEST, a powerful query processor for time series databases. TQuEST supports a novel but very useful class of queries which we call threshold queries. Threshold queries enable searches for time series whose values are above a user defined threshold at certain time intervals. Example queries are "report all ozone curves which are above their daily mean value at the same time as a given temperature curve exceeds 28 • C" or "report all blood value curves from patients whose values exceed a certain threshold one hour after the new medication was taken". TQuEST is based on a novel representation of time series which allows the query processor to access only the relevant parts of the time series. This enables an efficient execution of threshold queries. In particular, queries can be readjusted with interactive response times.