Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

Recent work has demonstrated that deep neural networks are vulnerable to adversarial examples---inputs that are almost indistinguishable from natural data and yet classified incorrectly by the network. In fact, some of the latest findings suggest that the existence of adversarial attacks may be an inherent weakness of deep learning models. To address this problem, we study the adversarial robustness of neural networks through the lens of robust optimization. This approach provides us with a broad and unifying view on much of the prior work on this topic. Its principled nature also enables us to identify methods for both training and attacking neural networks that are reliable and, in a certain sense, universal. In particular, they specify a concrete security guarantee that would protect against any adversary. These methods let us train networks with significantly improved resistance to a wide range of adversarial attacks. They also suggest the notion of security against a first-order adversary as a natural and broad security guarantee. We believe that robustness against such well-defined classes of adversaries is an important stepping stone towards fully resistant deep learning models. Code and pre-trained models are available at this https URL and this https URL.

Towards Deep Learning Models Resistant to Adversarial Attacks

Most existing machine learning classifiers are highly vulnerable to adversarial examples. An adversarial example is a sample of input data which has been modified very slightly in a way that is intended to cause a machine learning classifier to misclassify it. In many cases, these modifications can be so subtle that a human observer does not even notice the modification at all, yet the classifier still makes a mistake. Adversarial examples pose security concerns because they could be used to perform an attack on machine learning systems, even if the adversary has no access to the underlying model. Up to now, all previous work have assumed a threat model in which the adversary can feed data directly into the machine learning classifier. This is not always the case for systems operating in the physical world, for example those which are using signals from cameras and other sensors as an input. This paper shows that even in such physical world scenarios, machine learning systems are vulnerable to adversarial examples. We demonstrate this by feeding adversarial images obtained from cell-phone camera to an ImageNet Inception classifier and measuring the classification accuracy of the system. We find that a large fraction of adversarial examples are classified incorrectly even when perceived through the camera.

Adversarial examples in the physical world

Understanding and interpreting classification decisions of automated image classification systems is of high value in many applications, as it allows to verify the reasoning of the system and provides additional information to the human expert. Although machine learning methods are solving very successfully a plethora of tasks, they have in most cases the disadvantage of acting as a black box, not providing any information about what made them arrive at a particular decision. This work proposes a general solution to the problem of understanding classification decisions by pixel-wise decomposition of nonlinear classifiers. We introduce a methodology that allows to visualize the contributions of single pixels to predictions for kernel-based classifiers over Bag of Words features and for multilayered neural networks. These pixel contributions can be visualized as heatmaps and are provided to a human expert who can intuitively not only verify the validity of the classification decision, but also focus further analysis on regions of potential interest. We evaluate our method for classifiers trained on PASCAL VOC 2009 images, synthetic image data containing geometric shapes, the MNIST handwritten digits data set and for the pre-trained ImageNet model available as part of the Caffe open source package.

/pdf/on-pixel-wise-explanations-for-non-linear-classifier-4j59r9tlxm.pdf

On Pixel-Wise Explanations for Non-Linear Classifier Decisions by Layer-Wise Relevance Propagation.

The growing amount of web-based attacks poses a severe threat to the security of web applications. Signature-based detection techniques increasingly fail to cope with the variety and complexity of novel attack instances. As a remedy, we introduce a protocol-aware reverse HTTP proxy TokDoc (the token doctor), which intercepts requests and decides on a per-token basis whether a token requires automatic "healing". In particular, we propose an intelligent mangling technique, which, based on the decision of previously trained anomaly detectors, replaces suspicious parts in requests by benign data the system has seen in the past. Evaluation of our system in terms of accuracy is performed on two real-world data sets and a large variety of recent attacks. In comparison to state-of-the-art anomaly detectors, TokDoc is not only capable of detecting most attacks, but also significantly outperforms the other methods in terms of false positives. Runtime measurements show that our implementation can be deployed as an inline intrusion prevention system.

/pdf/tokdoc-a-self-healing-web-application-firewall-5a3oa074m0.pdf

TokDoc: a self-healing web application firewall

The article presents a general view of a class of decomposition algorithms for training Support Vector Machines (SVM) which are motivated by the method of feasible directions The first such algorithm for the pattern recognition SVM has been proposed in Joachims, T (1999, Scholkopf et al (Eds) Advances in kernel methods-Support vector learning (pp 185–208) MIT Press) Its extension to the regression SVM—the maximal inconsistency algorithm—has been recently presented by the author (Laskov, 2000, Solla, Leen, & Muller (Eds) Advances in neural information processing systems 12 (pp 484–490) MIT Press) A detailed account of both algorithms is carried out, complemented by theoretical investigation of the relationship between the two algorithms It is proved that the two algorithms are equivalent for the pattern recognition SVM, and the feasible direction interpretation of the maximal inconsistency algorithm is given for the regression SVM The experimental results demonstrate an order of magnitude decrease of training time in comparison with training without decomposition, and, most importantly, provide experimental evidence of the linear convergence rate of the feasible direction decomposition algorithms

/pdf/feasible-direction-decomposition-algorithms-for-training-24jpylwegd.pdf

Feasible Direction Decomposition Algorithms for Training Support Vector Machines

Current Voice-over-IP infrastructures lack defenses against unexpected network threats, such as zero-day exploits and computer worms. The possibility of such threats originates from the ongoing convergence of telecommunication and IP network infrastructures. As a countermeasure, we propose a self-learning system for detection of unknown and novel attacks in the Session Initiation Protocol (SIP). The system identifies anomalous content by embedding SIP messages to a feature space and determining deviation from a model of normality. The system adapts to network changes by automatically retraining itself while being hardened against targeted manipulations. Experiments conducted with realistic SIP traffic demonstrate the high detection performance of the proposed system at low false-positive rates.

/pdf/a-self-learning-system-for-detection-of-anomalous-sip-42m4jy3i37.pdf

A Self-learning System for Detection of Anomalous SIP Messages

A frequent problem in anomaly detection is to decide among different feature sets to be used. For example, various features are known in network intrusion detection based on packet headers, content byte streams or application level protocol parsing. A method for automatic feature selection in anomaly detection is proposed which determines optimal mixture coefficients for various sets of features. The method generalizes the support vector data description (SVDD) and can be expressed as a semi-infinite linear program that can be solved with standard techniques. The case of a single feature set can be handled as a particular case of the proposed method. The experimental evaluation of the new method on unsanitized HTTP data demonstrates that detectors using automatically selected features attain competitive performance, while sparing practitioners from a priori decisions on feature sets to be used.

Automatic feature selection for anomaly detection

The invention is concerned with a method and an apparatus for automatic comparison of at least two data sequences characterized in: an evaluation of a local relationship between any pair of subsequences in two or more sequences; an evaluation of a global relationship by means of aggregation of the evaluations of said local relationships.

Pavel Laskov

Papers

TokDoc: a self-healing web application firewall

Feasible Direction Decomposition Algorithms for Training Support Vector Machines

A Self-learning System for Detection of Anomalous SIP Messages

Automatic feature selection for anomaly detection

A method and apparatus for automatic comparison of data sequences