Some of them are related to the data distribution, notably the class imbalance of the training set (many more genuine transactions than fraudulent ones), the non-stationarity of the phenomenon (due to changes in the behavior of customers as well as in fraudsters), the large dimensionality and the overlapping classes (while fraudsters try to emulate cardholders behavior, genuine behaviors of cardholders might look strange or anomalous). The labeling process is constrained, as every day human investigators may contact only a small number of cardholders associated with the riskiest transactions and obtain the class (fraud or genuine) of the related transactions. The high cost of human labour, for assessing the transaction labels, leads to the labeling bottleneck [2]. In this context, an automatic Fraud Detection System (FDS) should support the activity of the investigators by letting them focus on the transactions with the highest fraud probability. From the perspective of the transactional service company, this is crucial in order to reduce the costs of the investigation activity and to retain the customer confidence. From a machine learning perspective it is important to keep an adequate balance between exploitation and exploration, i.e. between the short-term needs of providing good alerts to investigators, and the long-term goal of maintaining a high accuracy of the system (e.g. in the presence of concept drift). The issue of labeling the most informative data by minimizing the cost has been extensively addressed by active learning which can be considered as a specific instance of semi-supervised learning [8, 41], the domain studying how unlabeled and labeled data can both contribute to 2 Fabrizio Carcillo et al.

Ontology alignment is widely-used to find the correspondences between different ontologies in diverse fields.After discovering the alignments,several performance scores are available to evaluate them.The scores typically require the identified alignment and a reference containing the underlying actual correspondences of the given ontologies.The current trend in the alignment evaluation is to put forward a new score(e.g., precision, weighted precision, etc.)and to compare various alignments by juxtaposing the obtained scores. However,it is substantially provocative to select one measure among others for comparison.On top of that, claiming if one system has a better performance than one another cannot be substantiated solely by comparing two scalars.In this paper,we propose the statistical procedures which enable us to theoretically favor one system over one another.The McNemar's test is the statistical means by which the comparison of two ontology alignment systems over one matching task is drawn.The test applies to a 2x2 contingency table which can be constructed in two different ways based on the alignments,each of which has their own merits/pitfalls.The ways of the contingency table construction and various apposite statistics from the McNemar's test are elaborated in minute detail.In the case of having more than two alignment systems for comparison, the family-wise error rate is expected to happen. Thus, the ways of preventing such an error are also discussed.A directed graph visualizes the outcome of the McNemar's test in the presence of multiple alignment systems.From this graph, it is readily understood if one system is better than one another or if their differences are imperceptible.The proposed statistical methodologies are applied to the systems participated in the OAEI 2016 anatomy track, and also compares several well-known similarity metrics for the same matching problem.

The family of image visibility graphs (IVGs) have been recently introduced as simple algorithms by which scalar fields can be mapped into graphs. Here we explore the usefulness of such operator in the scenario of image processing and image classification. We demonstrate that the link architecture of the image visibility graphs encapsulates relevant information on the structure of the images and we explore their potential as image filters and compressors. We introduce several graph features, including the novel concept of Visibility Patches, and show through several examples that these features are highly informative, computationally efficient and universally applicable for general pattern recognition and image classification tasks.

A natural way of handling imbalanced data is to attempt to equalise the class frequencies and train the classifier of choice on balanced data. For two-class imbalanced problems, the classification success is typically measured by the geometric mean (GM) of the true positive and true negative rates. Here we prove that GM can be improved upon by instance selection, and give the theoretical conditions for such an improvement. We demonstrate that GM is non-monotonic with respect to the number of retained instances, which discourages systematic instance selection. We also show that balancing the distribution frequencies is inferior to a direct maximisation of GM. To verify our theoretical findings, we carried out an experimental study of 12 instance selection methods for imbalanced data, using 66 standard benchmark data sets. The results reveal possible room for new instance selection methods for imbalanced data.

Automatic detection of anomalies in space- and time-varying measurements is an important tool in several fields, e.g., fraud detection, climate analysis, or healthcare monitoring. We present an algorithm for detecting anomalous regions in multivariate spatio-temporal time-series, which allows for spotting the interesting parts in large amounts of data, including video and text data. In opposition to existing techniques for detecting isolated anomalous data points, we propose the "Maximally Divergent Intervals" (MDI) framework for unsupervised detection of coherent spatial regions and time intervals characterized by a high Kullback-Leibler divergence compared with all other data given. In this regard, we define an unbiased Kullback-Leibler divergence that allows for ranking regions of different size and show how to enable the algorithm to run on large-scale data sets in reasonable time using an interval proposal technique. Experiments on both synthetic and real data from various domains, such as climate analysis, video surveillance, and text forensics, demonstrate that our method is widely applicable and a valuable tool for finding interesting events in different types of data.

The first step in constructing a machine learning model is defining the features of the data set that can be used for optimal learning. In this work we discuss feature selection methods, which can be used to build better models, as well as achieve model interpretability. We applied these methods in the context of stress hotspot classification problem, to determine what microstructural characteristics can cause stress to build up in certain grains during uniaxial tensile deformation. The results show how some feature selection techniques are biased and demonstrate a preferred technique to get feature rankings for physical interpretations.

However, this variation is often aligned with author attributes such as age, gender, and geography, as well as more readily-available social network metadata. In this paper, we report new evidence on the link between language and social networks in the task of part-of-speech tagging. We find that tagger error rates are correlated with network structure, with high accuracy in some parts of the network, and lower accuracy elsewhere. As a result, tagger accuracy depends on training from a balanced sample of the network, rather than training on texts from a narrow subcommunity. We also describe our attempts to add robustness to stylistic variation, by building a mixture-of-experts model in which each expert is associated with a region of the social network. While prior work found that similar approaches yield performance improvements in sentiment analysis and entity linking, we were unable to obtain performance improvements in part-of-speech tagging, despite strong evidence for the link between part-of-speech error rates and social network structure.

One of the most common tasks performed by data scientists and data analysts are prediction and machine learning. This course will cover the basic components of building and applying prediction functions with an emphasis on practical applications. The course will provide basic grounding in concepts such as training and tests sets, overfitting, and error rates. The course will also introduce a range of model based and algorithmic machine learning methods including regression, classification trees, Naive Bayes, and random forests. The course will cover the complete process of building prediction functions including data collection, feature creation, algorithms, and evaluation.

A group of computer scientists have built a machine learning algorithm that can sniff out fake profiles lurking on social networks. It's likely that you and your Facebook friends have the same mutual friends. And on Twitter, it's also probable that your followers also follow the same people you do too due to common interests. These associations in social networks can be modeled on a graph as edges, where users are the vertices or nodes. The researchers from the Ben-Gurion University of the Negev, Israel, and the University of Washington, United States, hunted for fake social media accounts by developing an unsupervised learning algorithm that measures the probability of an edge existing between vertices.

An ensemble technique is characterized by the mechanism that generates the components and by the mechanism that combines them. A common way to achieve the consensus is to enable each component to equally participate in the aggregation process. A problem with this approach is that poor components are likely to negatively affect the quality of the consensus result. To address this issue, alternatives have been explored in the literature to build selective classifier and cluster ensembles, where only a subset of the components contributes to the computation of the consensus. Of the family of ensemble methods, outlier ensembles are the least studied. Only recently, the selection problem for outlier ensembles has been discussed. In this work we define a new graph-based class of ranking selection methods. A method in this class is characterized by two main steps: (1) Mapping the rankings onto a graph structure; and (2) Mining the resulting graph to identify a subset of rankings. We define a specific instance of the graph-based ranking selection class. Specifically, we map the problem of selecting ensemble components onto a mining problem in a graph. An extensive evaluation was conducted on a variety of heterogeneous data and methods. Our empirical results show that our approach outperforms state-of-the-art selective outlier ensemble techniques.