We develop and demonstrate a probabilistic method for classifying rare objects in surveys with the particular goal of building very pure samples. It works by modifying the output probabilities from a classifier so as to accommodate our expectation (priors) concerning the relative frequencies of different classes of objects. We demonstrate our method using the Discrete Source Classifier, a supervised classifier currently based on Support Vector Machines, which we are developing in preparation for the Gaia data analysis. DSC classifies objects using their very low resolution optical spectra. We look in detail at the problem of quasar classification, because identification of a pure quasar sample is necessary to define the Gaia astrometric reference frame. By varying a posterior probability threshold in DSC we can trade off sample completeness and contamination. We show, using our simulated data, that it is possible to achieve a pure sample of quasars (upper limit on contamination of 1 in 40,000) with a completeness of 65% at magnitudes of G=18.5, and 50% at G=20.0, even when quasars have a frequency of only 1 in every 2000 objects. The star sample completeness is simultaneously 99% with a contamination of 0.7%. Including parallax and proper motion in the classifier barely changes the results. We further show that not accounting for class priors in the target population leads to serious misclassifications and poor predictions for sample completeness and contamination. (Truncated)

Machine learning techniques are gaining prevalence in the production of a wide range of classifiers for complex real-world applications with nonuniform testing and misclassification costs. The increasing complexity of these applications poses a real challenge to resource management during learning and classification. In this work we introduce ACT (anytime cost-sensitive tree learner), a novel framework for operating in such complex environments. ACT is an anytime algorithm that allows learning time to be increased in return for lower classification costs. It builds a tree top-down and exploits additional time resources to obtain better estimations for the utility of the different candidate splits. Using sampling techniques, ACT approximates the cost of the subtree under each candidate split and favors the one with a minimal cost. As a stochastic algorithm, ACT is expected to be able to escape local minima, into which greedy methods may be trapped. Experiments with a variety of datasets were conducted to compare ACT to the state-of-the-art cost-sensitive tree learners. The results show that for the majority of domains ACT produces significantly less costly trees. ACT also exhibits good anytime behavior with diminishing returns.

It has been widely observed that there is no single "dominant" SAT solver; instead, different solvers perform best on different instances. Rather than following the traditional approach of choosing the best solver for a given class of instances, we advocate making this decision online on a per-instance basis. Building on previous work, we describe SATzilla, an automated approach for constructing per-instance algorithm portfolios for SAT that use so-called empirical hardness models to choose among their constituent solvers. This approach takes as input a distribution of problem instances and a set of component solvers, and constructs a portfolio optimizing a given objective function (such as mean runtime, percent of instances solved, or score in a competition). The excellent performance of SATzilla was independently verified in the 2007 SAT Competition, where our SATzilla07 solvers won three gold, one silver and one bronze medal. In this article, we go well beyond SATzilla07 by making the portfolio construction scalable and completely automated, and improving it by integrating local search solvers as candidate solvers, by predicting performance score instead of runtime, and by using hierarchical hardness models that take into account different types of SAT instances. We demonstrate the effectiveness of these new techniques in extensive experimental results on data sets including instances from the most recent SAT competition.

Art historians and archaeologists have long grappled with the regional classification of ancient Near Eastern ivory carvings. Based on the visual similarity of sculptures, individuals within these fields have proposed object assemblages linked to hypothesized regional production centers. Using quantitative rather than visual methods, we here approach this classification task by exploiting computational methods from machine learning currently used with success in a variety of statistical problems in science and engineering. We first construct a prediction function using 66 categorical features as inputs and regional style as output. The model assigns regional style group (RSG), with 98 percent prediction accuracy. We then rank these features by their mutual information with RSG, quantifying single-feature predictive power. Using the highest- ranking features in combination with nomographic visualization, we have found previously unknown relationships that may aid in the regional classification of these ivories and their interpretation in art historical context.

In order for agents to act on behalf of users, they will have to retrieve and integrate vast amounts of textual data on the World Wide Web. However, much of the useful data on the Web is neither grammatical nor formally structured, making querying difficult. Examples of these types of data sources are online classifieds like Craigslist and auction item listings like eBay. We call this unstructured, ungrammatical data "posts." The unstructured nature of posts makes query and integration difficult because the attributes are embedded within the text. Also, these attributes do not conform to standardized values, which prevents queries based on a common attribute value. The schema is unknown and the values may vary dramatically making accurate search difficult. Creating relational data for easy querying requires that we define a schema for the embedded attributes and extract values from the posts while standardizing these values. Traditional information extraction (IE) is inadequate to perform this task because it relies on clues from the data, such as structure or natural language, neither of which are found in posts. Furthermore, traditional information extraction does not incorporate data cleaning, which is necessary to accurately query and integrate the source. The two-step approach described in this paper creates relational data sets from unstructured and ungrammatical text by addressing both issues. To do this, we require a set of known entities called a "reference set." The first step aligns each post to each member of each reference set. This allows our algorithm to define a schema over the post and include standard values for the attributes defined by this schema. The second step performs information extraction for the attributes, including attributes not easily represented by reference sets, such as a price. In this manner we create a relational structure over previously unstructured data, supporting deep and accurate queries over the data as well as standard values for integration. Our experimental results show that our technique matches the posts to the reference set accurately and efficiently and outperforms state-of-the-art extraction systems on the extraction task from posts.

In this paper, we implement an anomaly detection system using the Dempster-Shafer method. Using two standard benchmark problems we show that by combining multiple signals it is possible to achieve better results than by using a single signal. We further show that by applying this approach to a real-world email dataset the algorithm works for email worm detection. Dempster-Shafer can be a promising method for anomaly detection problems with multiple features (data sources), and two or more classes.

Mining association rules is an important technique for discovering meaningful patterns in transaction databases. Many different measures of interestingness have been proposed for association rules. However, these measures fail to take the probabilistic properties of the mined data into account. In this paper, we start with presenting a simple probabilistic framework for transaction data which can be used to simulate transaction data when no associations are present. We use such data and a real-world database from a grocery outlet to explore the behavior of confidence and lift, two popular interest measures used for rule mining. The results show that confidence is systematically influenced by the frequency of the items in the left hand side of rules and that lift performs poorly to filter random noise in transaction data. Based on the probabilistic framework we develop two new interest measures, hyper-lift and hyper-confidence, which can be used to filter or order mined association rules. The new measures show significantly better performance than lift for applications where spurious rules are problematic.

Gesture is a non-verbal modality that can contribute crucial information to the understanding of natural language. But not all gestures are informative, and non-communicative hand motions may confuse natural language processing (NLP) and impede learning. People have little difficulty ignoring irrelevant hand movements and focusing on meaningful gestures, suggesting that an automatic system could also be trained to perform this task. However, the informativeness of a gesture is context-dependent and labeling enough data to cover all cases would be expensive. We present conditional modality fusion, a conditional hidden-variable model that learns to predict which gestures are salient for coreference resolution, the task of determining whether two noun phrases refer to the same semantic entity. Moreover, our approach uses only coreference annotations, and not annotations of gesture salience itself. We show that gesture features improve performance on coreference resolution, and that by attending only to gestures that are salient, our method achieves further significant gains. In addition, we show that the model of gesture salience learned in the context of coreference accords with human intuition, by demonstrating that gestures judged to be salient by our model can be used successfully to create multimedia keyframe summaries of video. These summaries are similar to those created by human raters, and significantly outperform summaries produced by baselines from the literature.

We study the efficiency of V-fold cross-validation (VFCV) for model selection from the non-asymptotic viewpoint, and suggest an improvement on it, which we call ``V-fold penalization''. Considering a particular (though simple) regression problem, we prove that VFCV with a bounded V is suboptimal for model selection, because it ``overpenalizes'' all the more that V is large. Hence, asymptotic optimality requires V to go to infinity. However, when the signal-to-noise ratio is low, it appears that overpenalizing is necessary, so that the optimal V is not always the larger one, despite of the variability issue. This is confirmed by some simulated data. In order to improve on the prediction performance of VFCV, we define a new model selection procedure, called ``V-fold penalization'' (penVF). It is a V-fold subsampling version of Efron's bootstrap penalties, so that it has the same computational cost as VFCV, while being more flexible. In a heteroscedastic regression framework, assuming the models to have a particular structure, we prove that penVF satisfies a non-asymptotic oracle inequality with a leading constant that tends to 1 when the sample size goes to infinity. In particular, this implies adaptivity to the smoothness of the regression function, even with a highly heteroscedastic noise. Moreover, it is easy to overpenalize with penVF, independently from the V parameter. A simulation study shows that this results in a significant improvement on VFCV in non-asymptotic situations.

We present a method for binary online classification of triggered but temporally blurred events that are embedded in noisy time series in the context of online discrimination between left and right imaginary hand-movement. In particular the goal of the binary classification problem is to obtain the decision, as fast and as reliably as possible from the recorded EEG single trials. To provide a probabilistic decision at every time-point t the presented method gathers information from two distinct sequences of features across time. In order to incorporate decisions from prior time-points we suggest an appropriate weighting scheme, that emphasizes time instances, providing a higher discriminatory power between the instantaneous class distributions of each feature, where the discriminatory power is quantified in terms of the Bayes error of misclassification. The effectiveness of this procedure is verified by its successful application in the 3rd BCI competition. Disclosure of the data after the competition revealed this approach to be superior with single trial error rates as low as 10.7, 11.5 and 16.7% for the three different subjects under study.