Novelty Detection in Projected Spaces for Structural Health Monitoring.

Benchmarking dynamic time warping for music retrieval.
We study the performance of three dynamic programming methods on music retrieval. The methods are designed for time series matching but can be directly applied to retrieval of music. Dynamic Time Warping (DTW) identifies an optimal alignment between two time series, and computes the matching cost corresponding to that alignment. Significant speed-ups can be achieved by constrained Dynamic Time Warping (cDTW), which narrows down the set of positions in one time series that can be matched with specific positions in the other time series. Both methods are designed for full sequence matching but can also be applied for subsequence matching, by using a sliding window over each database sequence to compute a matching score for each database subsequence. In addition, SPRING is a dynamic programming approach designed for subsequence matching, where the query is matched with a database subsequence without requiring the match length to be equal to the query length. SPRING has a lower computational cost than DTW and cDTW. Our database consists of a set of MIDI files taken from the web. Each MIDI file has been converted to a 2-dimensional time series, taking into account both note pitches and durations. We have used synthetic queries of fixed size and different noise levels. Surprisingly, when looking for the top-K best matches, all three approaches show similar behavior in terms of retrieval accuracy for small values of K. This suggests that for the specific application area, a computationally cheaper method, such as SPRING, is sufficient to retrieve the best top-K matches.
Tracking your steps on the track body sensor recordings of a controlled walking experiment.
Monitoring human motion has recently received great attention and can be used in many applications, such as human motion prediction. We present the collected data set from a body sensor network attached to the human body. The set of sensors consists of accelerometers measuring acceleration in three directions that are attached to the upper and lower back as well as the knees and ankles. In addition, pressures on the insoles are measured with four pressure sensors inside each shoe. Two types of motion are considered: walking backwards on a straight line and walking forwards on a figure-8 path. Finally, we study and present basic statistics of the data.
Preservation of Statistically Significant Patterns in Multiresolution 0%2D1 Data.

Automatic detection of onset and cessation of tree stem radius increase using dendrometer data.

Feature Extraction and Selection from Vibration Measurements for Structural Health Monitoring.

Feature Extraction and Selection from Vibration Measurements for Structural Health Monitoring.

Smoothed Prediction of the Onset of Tree Stem Radius Increase Based on Temperature Patterns.

Selection of important input variables for RBF network using partial derivatives.
Abstract. In regression problems, making accurate predictions is often the primary goal. Also, relevance of inputs in the prediction of an output would be valuable information in many cases. A sequential input selection algorithm for Radial basis function (SISAL-RBF) networks is presented to analyze importances of the inputs. The ranking of inputs is based on values, which are evaluated from the partial derivatives of the network. The proposed method is applied to benchmark data sets. It yields accurate prediction models, which are parsimonious in terms of the input variables. 1
Spatio%2Dtemporal Road Condition Forecasting with Markov Chains and Artificial Neural Networks.

Model Selection and Estimation Via Subjective User Preferences.

Segmenting Multi%2Dattribute Sequences Using Dynamic Bayesian Networks.
Discovering dependencies between attributes in multiattribute event sequences (multi-sequences), also known as synchronized multi-stream sequences, is an important problem in many domains, including monitoring systems and molecular biology. Many real-life multi-sequences have a segmental structure, with segments of differing complexities of attribute dependencies, which reflects a changing nature of the dependencies over time and space. In this paper we propose a new approach for discovering dependencies in multi-sequences which considers a possible segmental nature of such dependencies and tries to describe the multisequences in probabilistic terms using Dynamic Bayesian Networks (DBN). To accurately quantify such changing dependencies, we segment the multi-sequence by fitting an optimal DBN for each segment. We use the Bayesian Information Criterion (BIC) to select an optimal DBN structure and the number of segments of the multi-sequence. 1.
Compact and Understandable Descriptions of Mixtures of Bernoulli Distributions.
Abstract. Finite mixture models can be used in estimating complex, unknown probability distributions and also in clustering data. The parameters of the models form a complex representation and are not suitable for interpretation purposes as such. In this paper, we present a methodology to describe the finite mixture of multivariate Bernoulli distributions with a compact and understandable description. First, we cluster the data with the mixture model and subsequently extract the maximal frequent itemsets from the cluster-specific data sets. The mixture model is used to model the data set globally and the frequent itemsets model the marginal distributions of the partitioned data locally. We present the results in understandable terms that reflect the domain properties of the data. In our application of analyzing DNA copy number amplifications, the descriptions of amplification patterns are represented in nomenclature used in literature to report amplification patterns and generally used by domain experts in biology and medicine. 1
Mixture Modeling of DNA Copy Number Amplification Patterns in Cancer.
amplifications collected from more than 4500 neoplasm cases. Based on the 0-1 representation of the data, we trained finite mixtures of multivariate Bernoulli distributions using the EM algorithm to describe the inherent structure in the data. The resulting component distributions of the mixtures of Bernoulli distributions yielded plausible and localized amplification patterns. Individual amplification patterns were tested for their role in cancer groups formed with known risk associations. Our detailed analysis of chromosome 1 showed that asbestos-exposure related and hormonal imbalance-associated cancers were clustered and specific chromosome bands, 1p34 and 1q42, were identified. These sites contain cancer genes, which might explain the condition-specific selection of these loci for amplification. 1
Modeling the effects of varying data quality on trend detection in environmental monitoring.

Analysis of Linux Evolution Using Aligned Source Code Segments.

Analysis of Fast Input Selection Application in Time Series Prediction.%
Abstract. In time series prediction, accuracy of predictions is often the primary goal. At the same time, however, it would be very desirable if we could give interpretation to the system under study. For this goal, we have devised a fast input selection algorithm to choose a parsimonious, or sparse set of input variables. The method is an algorithm in the spirit of backward selection used in conjunction with the resampling procedure. In this paper, our strategy is to select a sparse set of inputs using linear models and after that the selected inputs are also used in the nonlinear prediction based on multi-layer perceptron networks. We compare the prediction accuracy of our parsimonious non-linear models with the linear models and the regularized non-linear perceptron networks. Furthermore, we quantify the importance of the individual input variables in the non-linear models using the partial derivatives. The experiments in a problem of electricity load prediction demonstrate that the fast input selection method yields accurate and parsimonious prediction models giving insight to the original problem. 1
Analysis of Fast Input Selection Application in Time Series Prediction.%
Abstract. In time series prediction, accuracy of predictions is often the primary goal. At the same time, however, it would be very desirable if we could give interpretation to the system under study. For this goal, we have devised a fast input selection algorithm to choose a parsimonious, or sparse set of input variables. The method is an algorithm in the spirit of backward selection used in conjunction with the resampling procedure. In this paper, our strategy is to select a sparse set of inputs using linear models and after that the selected inputs are also used in the nonlinear prediction based on multi-layer perceptron networks. We compare the prediction accuracy of our parsimonious non-linear models with the linear models and the regularized non-linear perceptron networks. Furthermore, we quantify the importance of the individual input variables in the non-linear models using the partial derivatives. The experiments in a problem of electricity load prediction demonstrate that the fast input selection method yields accurate and parsimonious prediction models giving insight to the original problem. 1
Combining Measurement Quality into Monitoring Trends in Foliar Nutrient Concentrations.

Data integration and visualization system for enabling conceptual biology.

Input Selection for Long%2DTerm Prediction of Time Series.
Abstract. Prediction of time series is an important problem in many areas of science and engineering. Extending the horizon of predictions further to the future is the challenging and difficult task of long-term prediction. In this paper, we investigate the problem of selecting noncontiguous input variables for an autoregressive prediction model in order to improve the prediction ability. We present an algorithm in the spirit of backward selection which removes variables sequentially from the prediction models based on the significance of the individual regressors. We successfully test the algorithm with a non-linear system by selecting inputs with a linear model and finally train a non-linear predictor with the selected variables on Santa Fe laser data set. 1
Model%2DIndependent Bounding of the Supports of Boolean Formulae in Binary Data.
Data mining algorithms such as the Apriori method for finding frequent sets in sparse binary data can be used for efficient computation of a large number of summaries from huge data sets. The collection of frequent sets gives a collection of marginal frequencies about the underlying data set. Sometimes, we would like to use a collection of such marginal frequencies instead of the entire data set (e.g. when the original data is inaccessible for confidentiality reasons) to compute other interesting summaries. Using combinatorial arguments, we may obtain tight upper and lower bounds on the values of inferred summaries. In this paper, we consider a class of summaries wider than frequent sets, namely frequencies of arbitrary Boolean formulae. Given frequencies of a number of any different Boolean formulae, we consider the problem of finding tight bounds on the frequency of another arbitrary formula. We give a general formulation of the problem of bounding formula frequencies given some background information, and show how the bounds can be obtained by solving a linear programming problem. We illustrate the accuracy of the bounds by giving empirical results on real data sets.
Mixture Models and Frequent Sets Combining Global and Local Methods for 0%2D1 Data.

Image Analysis for Detecting Faulty Spots from Microarray Images.
Microarrays allow the monitoring of thousands of genes simultaneously. Before a measure of gene activity...
Model%2Dindependent Bounding of the Supports of Boolean Formulae in Binary Data.
Data mining algorithms such as the Apriori method for finding frequent sets in sparse binary data can be used for efficient computation of a large number of summaries from huge data sets. The collection of frequent sets gives a collection of marginal frequencies about the underlying data set. Sometimes, we would like to use a collection of such marginal frequencies instead of the entire data set (e.g. when the original data is inaccessible for confidentiality reasons) to compute other interesting summaries. Using combinatorial arguments, we may obtain tight upper and lower bounds on the values of inferred summaries. In this paper, we consider a class of summaries wider than frequent sets, namely frequencies of arbitrary Boolean formulae. Given frequencies of a number of any different Boolean formulae, we consider the problem of finding tight bounds on the frequency of another arbitrary formula. We give a general formulation of the problem of bounding formula frequencies given some background information, and show how the bounds can be obtained by solving a linear programming problem. We illustrate the accuracy of the bounds by giving empirical results on real data sets.
An Automated Report Generation Tool for the Data Understanding Phase.
To prepare and model data successfully, the data miner needs to be aware of the properties of the data manifold. In this paper, the outline of a tool for automatically generating data survey reports for this purpose is described. The report combines linguistic descriptions (rules) and statistical measures with visualizations. Together these provide both quantitative and qualitative information and help the user to form a mental model of the data. The main focus is on describing the cluster structure and the contents of the clusters. The data is clustered using a novel algorithm based on the Self-Organizing Map. The rules describing the clusters are selected using a signi cance measure based on the con dence on their characterizing and discriminating properties.
Quantization of Continuous Input Variables for Binary Classification.
Quantization of continuous variables is important in data analysis, especially for some model classes such as Bayesian networks and decision trees, which use discrete variables. Often, the discretization is based on the distribution of the input variables only whereas additional information, for example in form of class membership is frequently present and could be used to improve the quality of the results. In this paper, quantization methods based on equal width interval, maximum entropy, maximum mutual information and the novel approach based on maximum mutual information combined with entropy are considered. The two former approaches do not take the class membership into account whereas the two latter approaches do. The relative merits of each method are compared in an empirical setting, where results are shown for two data sets in a direct marketing problem, and the quality of quantization is measured by mutual information and the performance of Naive Bayes and C5 decision tree classifiers.
Call%2DBased Fraud Detection in Mobile Communication Networks Using a Hierarchical Regime%2DSwitching Model.
Fraud causes substantial losses to telecommunication carriers. Detection systems which automatically detect illegal use of the network can be used to alleviate the problem. Previous approaches worked on features derived from the call patterns of individual users. In this paper we present a call-based detection system based on a hierarchical regime-switching model. The detection problem is formulated as an inference problem on the regime probabilities. Inference is implemented by applying the junction tree algorithm to the underlying graphical model. The dynamics are learned from data using the EM algorithm and subsequent discriminative training. The methods are assessed using fraud data from a real mobile communication network. 1 INTRODUCTION Fraud is costly to a network carrier both in terms of lost income and wasted capacity. It has been estimated that the telecommunication industry looses approximately 2-5% of its total revenue to fraud. The true losses are expected to b...
