Keystroke dynamics is one of the most widely adopted identity verification techniques in remote systems. It is based on modeling users’ specific patterns of typing on the keyboard. When utilized in conjunction with the commonly used passwords, the use of keystroke dynamics can dramatically increase the level of security without interfering with the user experience. However, aspects of keystroke dynamics that applied on passwords, such as processing keystroke events and storing feature vectors or user models, can expose users to identity theft and a new set of privacy risks, thus questioning the added value of keystroke dynamics. In addition, common encryption techniques will be unable to mitigate these threats, since the user's behavior changes from one session to another. In this paper, we suggest key grouping as an obfuscation method to ensure keystroke dynamics privacy. When applied on the keystroke events, the key grouping dramatically reduces the possibility of password theft. To perform the key grouping optimally, we present a novel method which produces groups that can integrated with any keystroke dynamics algorithm. Our method divides the keys into groups using hierarchical clustering with dedicated statistical heuristics algorithm. We tested our method's key grouping output on five keystroke dynamics algorithms using a public dataset and managed to show a consistent improvement of up to 7% in the AUC over other, more intuitive key groupings and random key groupings.
No abstract is provided for this article.
The widespread adoption of machine learning (ML) techniques and the extensive expertise required to apply them have led to increased interest in automated ML solutions that reduce the need for human intervention. One of the main challenges in applying ML to previously unseen problems is algorithm selection – the identification of high-performing algorithm(s) for a given dataset, task, and evaluation measure. This study addresses the algorithm selection challenge for data clustering, a fundamental task in data mining that is aimed at grouping similar objects. We present MARCO-GE, a novel meta-learning approach for the automated recommendation of clustering algorithms. MARCO-GE first transforms datasets into graphs and then utilizes a graph convolutional neural network technique to extract their latent representation. Using the embedding representations obtained, MARCO-GE trains a ranking meta-model capable of accurately recommending top-performing algorithms for a new dataset and clustering evaluation measure. An extensive evaluation on 210 datasets, 17 clustering algorithms, and 10 clustering measures demonstrates the effectiveness of our approach and its superiority in terms of predictive and generalization performance over state-of-the-art clustering meta-learning approaches.
Despite continuous investments in data technologies, the latency of querying data still poses a significant challenge. Modern analytic solutions require near real-time responsiveness both to make them interactive and to support automated processing. Current technologies (Hadoop, Spark, Dataflow) scan the dataset to execute queries. They focus on providing a scalable data storage to maximize task execution speed. We argue that these solutions fail to offer an adequate level of interactivity since they depend on continual access to data. In this paper we present a method for query approximation, also known as approximate query processing (AQP), that reduce the need to scan data during inference (query calculation), thus enabling a rapid query processing tool. We use LSTM network to learn the relationship between queries and their results, and to provide a rapid inference layer for predicting query results. Our method (referred as ``Hunch``) produces a lightweight LSTM network which provides a high query throughput. We evaluated our method using twelve datasets and compared to state-of-the-art AQP engines (VerdictDB, BlinkDB) from query latency, model weight and accuracy perspectives. The results show that our method predicted queries' results with a normalized root mean squared error (NRMSE) ranging from approximately 1\% to 4\% which in the majority of our data sets was better then the compared benchmarks. Moreover, our method was able to predict up to 120,000 queries in a second (streamed together), and with a single query latency of no more than 2ms.
Drug-drug interactions are preventable causes of medical injuries and often result in doctor and emergency room visits. Computational techniques can be used to predict potential drug-drug interactions. We approach the drug-drug interaction prediction problem as a link prediction problem and present two novel methods for drug-drug interaction prediction based on artificial neural networks and factor propagation over graph nodes: adjacency matrix factorization (AMF) and adjacency matrix factorization with propagation (AMFP). We conduct a retrospective analysis by training our models on a previous release of the DrugBank database with 1,141 drugs and 45,296 drug-drug interactions and evaluate the results on a later version of DrugBank with 1,440 drugs and 248,146 drug-drug interactions. Additionally, we perform a holdout analysis using DrugBank. We report an area under the receiver operating characteristic curve score of 0.807 and 0.990 for the retrospective and holdout analyses respectively. Finally, we create an ensemble-based classifier using AMF, AMFP, and existing link prediction methods and obtain an area under the receiver operating characteristic curve of 0.814 and 0.991 for the retrospective and the holdout analyses. We demonstrate that AMF and AMFP provide state of the art results compared to existing methods and that the ensemble-based classifier improves the performance by combining various predictors. Additionally, we compare our methods with multi-source data-based predictors using cross-validation. In the multi-source data comparison, our methods outperform various ensembles created using 29 different predictors based on several data sources. These results suggest that AMF, AMFP, and the proposed ensemble-based classifier can provide important information during drug development and regarding drug prescription given only partial or noisy data. Additionally, the results indicate that the interaction network (known DDIs) is the most useful data source for identifying potential DDIs and that our methods take advantage of it better than the other methods investigated. The methods we present can also be used to solve other link prediction problems. Drug embeddings (compressed representations) created when training our models using the interaction network have been made public.
No abstract is provided for this article.
Driving under the influence of alcohol is a widespread phenomenon in the US where it is considered a major cause of fatal accidents. In this research we present a novel approach and concept for detecting intoxication from motion differences obtained by the sensors of wearable devices. We formalize the problem of drunkenness detection as a supervised machine learning task, both as a binary classification problem (drunk or sober) and a regression problem (the breath alcohol content level). In order to test our approach, we collected data from 30 different subjects (patrons at three bars) using Google Glass and the LG G-watch, Microsoft Band, and Samsung Galaxy S4. We validated our results against an admissible breathalyzer used by the police. A system based on this concept, successfully detected intoxication and achieved the following results: 0.95 AUC and 0.05 FPR, given a fixed TPR of 1.0. Applications based on our system can be used to analyze the free gait of drinkers when they walk from the car to the bar and vice-versa, in order to alert people, or even a connected car and prevent people from driving under the influence of alcohol.
One of the challenges that companies face when launching a campaign to promote new services is selecting the 'right' customers for the campaign, i.e., customers with the highest probability of a positive response. Active learning can be used to efficiently identify this set of customers. It can also prevent approach to non-relevant customers and reduce the campaign's cost. The problem is more challenging when parallel campaigns for multiple new services are launched, given a constraint on the number of promotions that can be offered to the same customer during a defined period of time. The goal is to maximize the total net profit. In this paper we present MutiCamp, a new cost sensitive active learning based algorithm that uses the Hungarian Algorithm to find the optimal match between campaigns and customers. MultiCamp was tested on a real world dataset using a decision tree classifier. Results were compared to a random baseline, indicating the superiority of the proposed algorithm.
Intrusion detection is the process of monitoring and analyzing the events occurring in a computer system in order to detect signs of security problems. The problem of intrusion detection can be solved using anomaly detection techniques. For instance, one is given a set of connection data belonging to different classes (normal activity, different attacks) and the aim is to construct a classifier that accurately classifies new unlabeled connections data. Clustering methods can be used to detect anomaly in data which might implies intrusion of a new type. This chapter gives a critical summary of anomaly detection research for intrusion detection. This chapter surveys a list of research projects that apply anomaly detection techniques to intrusion detection. Finally some directions for research are given.Request access from your librarian to read this chapter's full text.
No abstract is provided for this article.
Ensemble methods are considered the state‐of‐the art solution for many machine learning challenges. Such methods improve the predictive performance of a single model by training multiple models and combining their predictions. This paper introduce the concept of ensemble learning, reviews traditional, novel and state‐of‐the‐art ensemble methods and discusses current challenges and trends in the field. This article is categorized under: Algorithmic Development > Ensemble Methods Technologies > Machine Learning Technologies > Classification
The increasing usage of machine-learning models in critical domains has recently stressed the necessity of interpretable machine-learning models. In areas like healthcare, finary – the model consumer must understand the rationale behind the model output in order to use it when making a decision. For this reason, it is impossible to use black-box models in these scenarios, regardless of their high predictive performance. Decision forests, and in particular Gradient Boosting Decision Trees (GBDT), are examples of this kind of model. GBDT models are considered the state-of-the-art in many classification challenges, reflected by the fact that the majority of Kaggle’s recent winners used GBDT methods as a part of their solution (such as XGBoost). But despite their superior predictive performance, they cannot be used in tasks that require transparency. This paper presents a novel method for transforming a decision forest of any kind into an interpretable decision tree. The method extends the tool-set available for machine learning practitioners, who want to exploit the interpretability of decision trees without significantly impairing the predictive performance gained by GBDT models like XGBoost. We show in an empirical evaluation that in some cases the generated tree is able to approximate the predictive performance of a XGBoost model while enabling better transparency of the outputs.
W ikipedia, like other encyclopedias, includes biographies of notable people. However, because it is jointly written by many contributors, it is subject to constant manipulation by contributors attempting to add biographies of non‐notable people. Over time, W ikipedia has developed inclusion criteria for notable people (e.g., receiving a significant award) based on which newly contributed biographies are evaluated. In this paper we present and analyze a set of simple indicators that can be used to predict which article will eventually be accepted. These indicators do not refer to the content itself, but to meta‐content features (such as the number of categories that the biography is associated with) and to author‐based features (such as if it is a first‐time author). By training a classifier on these features, we successfully reached a high predictive performance (area under the receiver operating characteristic [ ROC ] curve [ AUC ] of 0.97) even though we overlooked the actual biography text.
Selecting the best classifier among the available ones is a difficult task, especially when only instances of one class exist. In this work we examine the notion of combining one-class classifiers as an alternative for selecting the best classifier. In particular, we propose two new one-class classification performance measures to weigh classifiers and show that a simple ensemble that implements these measures can outperform the most popular one-class ensembles. Furthermore, we propose a new one-class ensemble scheme, TUPSO, which uses meta-learning to combine one-class classifiers. Our experiments demonstrate the superiority of TUPSO over all other tested ensembles and show that the TUPSO performance is statistically indistinguishable from that of the hypothetical best classifier.
A group may appreciate recommendations on items that fit their joint preferences. When the members' actual preferences are unknown, a recommendation can be made with the aid of collaborative filtering methods. We offer to narrow down the recommended list of items by eliciting the users' actual preferences. Our final goal is to output top-k preferred items to the group out of the top-N recommendations provided by the recommender system (k<N), where one of the items is a necessary winner. We propose an iterative preference elicitation method, where users are required to provide item ratings per request. We suggest a heuristic that attempts to minimize the preference elicitation effort under two aggregation strategies. We evaluate our methods on real-world Netflix data as well as on simulated data which allows us to study different cases. We show that preference elicitation effort can be cut in up to 90% while preserving the most preferred items in the narrowed list.
Many applications that employ data mining techniques involve mining data that include private and sensitive information about the subjects. One way to enable effective data mining while preserving privacy is to anonymize the data set that includes private information about subjects before being released for data mining. One way to anonymize data set is to manipulate its content so that the records adhere to k-anonymity. Two common manipulation techniques used to achieve k-anonymity of a data set are generalization and suppression. Generalization refers to replacing a value with a less specific but semantically consistent value, while suppression refers to not releasing a value at all. Generalization is more commonly applied in this domain since suppression may dramatically reduce the quality of the data mining results if not properly used. However, generalization presents a major drawback as it requires a manually generated domain hierarchy taxonomy for every quasi-identifier in the data set on which k-anonymity has to be performed. In this paper, we propose a new method for achieving k-anonymity named K-anonymity of Classification Trees Using Suppression (kACTUS). In kACTUS, efficient multidimensional suppression is performed, i.e., values are suppressed only on certain records depending on other attribute values, without the need for manually produced domain hierarchy trees. Thus, in kACTUS, we identify attributes that have less influence on the classification of the data records and suppress them if needed in order to comply with k-anonymity. The kACTUS method was evaluated on 10 separate data sets to evaluate its accuracy as compared to other k-anonymity generalization- and suppression-based methods. Encouraging results suggest that kACTUS' predictive performance is better than that of existing k-anonymity algorithms. Specifically, on average, the accuracies of TDS, TDR, and kADET are lower than kACTUS in 3.5, 3.3, and 1.9 percent, respectively, despite their usage of manually defined domain trees. The accuracy gap is increased to 5.3, 4.3, and 3.1 percent, respectively, when no domain trees are used.
Anomaly detection is a well-known task that involves the identification of abnormal events that occur relatively infrequently. Methods for improving anomaly detection performance have been widely studied. However, no studies utilizing test-time augmentation (TTA) for anomaly detection in tabular data have been performed. TTA involves aggregating the predictions of several synthetic versions of a given test sample; TTA produces different points of view for a specific test instance and might decrease its prediction bias. We propose the Test-Time Augmentation for anomaly Detection (TTAD) technique, a TTA-based method aimed at improving anomaly detection performance. TTAD augments a test instance based on its nearest neighbors; various methods, including the k-Means centroid and SMOTE methods, are used to produce the augmentations. Our technique utilizes a Siamese network to learn an advanced distance metric when retrieving a test instance's neighbors. Our experiments show that the anomaly detector that uses our TTA technique achieved significantly higher AUC results on all datasets evaluated.
It is our great pleasure to welcome you to the 2013 ACM International Workshop on Computational Scientometrics: Theory and Applications -- CSTA'13. This is the first time when the CSTA workshop has been organized in the ACM International Conference on Information and Knowledge Management. The primary goals and objectives of the workshop are to promote both theoretical results and practical applications within digital libraries to better answer questions such as how do research ideas emerge, evolve, or disappear as a topic, what is a good measure of quality of published works, what are the most promising areas of research, how authors connect and influence each other, who are the experts in a field, what works are similar, and who funds a particular research topic. The workshop aims at bringing together researchers with diverse interdisciplinary backgrounds interested in mining the web, large digital libraries and other relevant databases for research related publications and data. The combination of classical bibliometrics and novel text mining provides a synergy unavailable within each approach taken independently. In this proposed workshop, we also focus on bibliometrics analysis by using sophisticated text mining, or natural language processing methods, which will enable researchers to generate innovative research topics, e.g. full-text citation analysis. We expect people interested in the practical applications within digital libraries such as citation analysis and recommendation, scientific and research trends, expert finding, and collaborator recommendation, to attend the CSTA'13 workshop. The call for papers attracted submissions from Europe, Australia, and the United States. The program committee accepted 6 papers that cover topics, including program committees recommendation for academic conferences, collaborators recommendation, and citations classification and labeling. In addition, the program includes a keynote speech by Xiaozhong Liu on Full-text Citation Analysis and its Application.
