End-to-end machine learning malware detectors are vulnerable to adversarial EXEmples, carefully-crafted malicious programs that evade detection through minimal perturbations. Such attacks typically operate by either replacing unused content (patch attacks), or injecting new patterns (content-injection attacks). To counter these attacks, recent research has focused on certification methods for end-to-end models that aim to prove robustness guarantees within bounded perturbation sizes. However, existing approaches (i) are not robust against content-injection manipulations and (ii) provide only probabilistic guarantees for perturbations that are negligible relative to the overall program size. Hence, in this paper we address these limitations through a novel deterministic certification schema based on (de)randomized smoothing. Our defense splits each executable into non-overlapping chunks and classifies them independently. The final decision is obtained via majority voting across all chunks, ensuring that localized modifications, such as injected or patched code, influence only a limited subset of chunks and have minimal impact in the overall classification. This design guarantees that each chunk either contains or does not contain an adversarial perturbation, enabling us to (i) handle manipulations occurring at arbitrary locations within the program, and (ii) compute deterministic estimates of the perturbation magnitude required to evade detection. We demonstrate the effectiveness of our certification schema through extensive experimental analysis, comparing our defense against a range of state-of-the-art attacks and defenses. The results show that our approach achieves unmatched robustness across all tested attack scenarios, substantially outperforming competing defenses.

The uses of machine learning (ML) have snowballed in recent years. In many cases, ML models are highly complex, and their operation is beyond the understanding of human decision-makers. Nevertheless, some uses of ML models involve high-stakes and safety-critical applications. Explainable artificial intelligence (XAI) aims to help human decision-makers in understanding the operation of such complex ML models, thus eliciting trust in their operation. Unfortunately, the majority of past XAI work is based on informal approaches, that offer no guarantees of rigor. Unsurprisingly, there exists comprehensive experimental and theoretical evidence confirming that informal methods of XAI can provide human-decision makers with erroneous information. Logic-based XAI represents a rigorous approach to explainability; it is model-based and offers the strongest guarantees of rigor of computed explanations. However, a well-known drawback of logic-based XAI is the complexity of logic reasoning, especially for highly complex ML models. Recent work proposed distance-restricted explanations, i.e. explanations that are rigorous provided the distance to a given input is small enough. Distance-restricted explainability is tightly related with adversarial robustness, and it has been shown to scale for moderately complex ML models, but the number of inputs still represents a key limiting factor. This paper investigates novel algorithms for scaling up the performance of logic-based explainers when computing and enumerating ML model explanations with a large number of inputs.

Abstract Given a machine learning (ML) model and a prediction, explanations can be defined as sets of features which are sufficient for the prediction. In some applications, and besides asking for an explanation, it is also critical to understand whether sensitive features can occur in some explanation, or whether a non-interesting feature must occur in all explanations. This paper starts by relating such queries respectively with the problems of relevancy and necessity in logic-based abduction. The paper then proves membership and hardness results for several families of ML classifiers. Afterwards the paper proposes concrete algorithms for two classes of classifiers. The experimental results confirm the scalability of the proposed algorithms.

Machine learning has become an appealing signature-less approach to detect and classify malware because of its ability to generalize to never-before-seen samples and to handle large volumes of data. While traditional feature-based approaches rely on the manual design of hand-crafted features based on experts’ knowledge of the domain, deep learning approaches replace the manual feature engineering process by an underlying system, typically consisting of a neural network with multiple layers, that perform both feature learning and classification altogether. However, the combination of both approaches could substantially enhance detection systems. In this paper we present an hybrid approach to address the task of malware classification by fusing multiple types of features defined by experts and features learned through deep learning from raw data. In particular, our approach relies on deep learning to extract N-gram like features from the assembly language instructions and the bytes of malware, and texture patterns and shapelet-based features from malware’s grayscale image representation and structural entropy, respectively. These deep features are later passed as input to a gradient boosting model that combines the deep features and the hand-crafted features using an early-fusion mechanism. The suitability of our approach has been evaluated on the Microsoft Malware Classification Challenge benchmark and results show that the proposed solution achieves state-of-the-art performance and outperforms gradient boosting and deep learning methods in the literature.

Malware detection and classification is a challenging problem and an active area of research. Traditional machine learning methods depend almost entirely on the ability to extract a set of discriminative features into which characterize malware. However, this feature engineering process is very time consuming. On the contrary, deep learning methods replace manual feature engineering by a system that performs both feature extraction and classification from raw data at once. Despite that, a major shortfall of these methods is their inhability to consider multiple disparate sources of information when performing classification, leading them to perform poorly when compared to multimodal approaches. In this work, we introduce Orthrus, a new bimodal approach to categorize malware into families based on deep learning. Orthrus combines two modalities of data: (1) the byte sequence representing the malware's binary content, and (2) the assembly language instructions extracted from the assembly language source code of malware, and performs automatic feature learning and classification with a convolutional neural network. The idea is to benefit from multiple feature types to reflect malware's characteristics. The experiments carried on the Microsoft Malware Classification Challenge dataset show that our proposed solution achieves higher classification performance than deep learning approaches in the literature and n-gram based methods.

The struggle between security analysts and malware developers is a never-ending battle with the complexity of malware changing as quickly as innovation grows. Current state-of-the-art research focus on the development and application of machine learning techniques for malware detection due to its ability to keep pace with malware evolution. This survey aims at providing a systematic and detailed overview of machine learning techniques for malware detection and in particular, deep learning techniques. The main contributions of the paper are: (1) it provides a complete description of the methods and features in a traditional machine learning workflow for malware detection and classification, (2) it explores the challenges and limitations of traditional machine learning and (3) it analyzes recent trends and developments in the field with special emphasis on deep learning approaches. Furthermore, (4) it presents the research issues and unsolved challenges of the state-of-the-art techniques and (5) it discusses the new directions of research. The survey helps researchers to have an understanding of the malware detection field and of the new developments and directions of research explored by the scientific community to tackle the problem.

Malware detection and classification is a challenging problem and an active area of research. Particular challenges include how to best treat and preprocess malicious executables in order to feed machine learning algorithms. Novel approaches in the literature treat an executable as a sequence of bytes or as a sequence of assembly language instructions. However, in those approaches the hierarchical structure of programs is not taken into consideration. An executable exhibits various levels of spatial correlation. Adjacent code instructions are correlated spatially but that is not necessarily the case. Function calls and jump commands transfer the control of the program to a different point in the instruction stream. Furthermore, these discontinuities are maintained when treating the binary as a sequence of byte values. In addition, functions might be arranged randomly if addresses are correctly reorganized. To address these issues we propose a Hierarchical Convolutional Network (HCN) for malware classification. It has two levels of convolutional blocks applied at the mnemonic-level and at the function-level, enabling us to extract n-gram like features from both levels when constructing the malware representation. We validate our HCN method on the dataset released for the Microsoft Malware Classification Challenge, outperforming almost every deep learning method in the literature.

The number of malicious programs has grown both in number and in sophistication. Analyzing the malicious intent of vast amounts of data requires huge resources and thus, effective categorization of malware is required. In this paper, the content of a malicious program is represented as an entropy stream, where each value describes the amount of entropy of a small chunk of code in a specific location of the file. Wavelet transforms are then applied to this entropy signal to describe the variation in the entropic energy. Motivated by the visual similarity between streams of entropy of malicious software belonging to the same family, we propose a file agnostic deep learning approach for categorization of malware. Our method exploits the fact that most variants are generated by using common obfuscation techniques and that compression and encryption algorithms retain some properties present in the original code. This allows us to find discriminative patterns that almost all variants in a family share. Our method has been evaluated using the data provided by Microsoft for the BigData Innovators Gathering Anti-Malware Prediction Challenge, and achieved promising results in comparison with the State of the Art.

Reducing energy consumption in large-scale computing facilities has become a major concern in recent years. The large number of computing nodes, resources heterogeneity and diversity of application requirements are factors that turn the scheduling into an NP-Hard problem. Evolutionary algorithms have proved to be effective for scheduling applications. In this paper, we present a novel approach combining particle swarm optimization and a genetic algorithm to solving the resource matching and scheduling of parallel applications in Federated cluster environments. The proposed hybrid meta-heuristic, referred to as MPSO-FGA, not only minimizes the overall energy consumption but also the makespan for a whole workload. The experimental results show the superiority of evolutionary algorithms over basic heuristics. The hybrid meta-heuristic is able to obtain similar results to a genetic algorithm in terms of energy consumption and makespan but reducing the time for scheduling decisions by two orders of magnitude.

Reducing energy consumption in large-scale computing facilities has become a major concern in recent years. Most of the techniques have focused on determining the computing requirements based on load predictions and thus turning unnecessary nodes on and off. Nevertheless, once the available resources have been configured, new opportunities arise for reducing energy consumption by providing optimal matching of parallel applications to the available computing nodes. Current research in scheduling has concentrated on not only optimizing the energy consumed by the processors but also optimizing the makespan, i.e., job completion time. The large number of heterogeneous computing nodes and variability of application-tasks are factors that make the scheduling an NP-Hard problem. Our aim in this paper is a multi-objective genetic algorithm based on a weighted blacklist able to generate scheduling decisions that globally optimizes the energy consumption and the makespan.

Minimum Satisfiability (MinSAT) denotes one of the optimization versions of the Boolean Satisfiability (SAT) problem. In some settings MinSAT is preferred to using Maximum Satis-fiability (MaxSAT). Several encodings and dedicated branch and bound algorithms for MinSAT have been recently proposed, and evaluated on small challenging randomly generated instances. Motivated by the observation that current best performing MaxSAT algorithms for structured and industrial instances are based on computing unsatisfiable cores with a SAT solver, this paper proposes novel approaches for MinSAT, that also target these instances. First, the paper proposes an algorithm based on iteratively calling a SAT solver which uses the computed models to relax clauses. Second, the paper proposes group-based MinSAT solving, which is essentially a novel reduction of the MinSAT problem into the Group MaxSAT problem. For a given MinSAT instance, the resulting Group MaxSAT formula is then translated into a standard MaxSAT formula which specifically targets unsatisfiability-based MaxSAT algorithms. Experimental results indicate that, similarly to MaxSAT, the proposed approaches outperform branch and bound algorithms on problem instances obtained from practical applications.

Many decision and optimization problems in electronic design automation (EDA) can be solved with Boolean satisfiability (SAT). Moreover, well-known extensions of SAT also find application in EDA, including pseudo-Boolean optimization, quantified Boolean formulas, multi-valued SAT and, more recently, Maximum Satisfiability (MaxSAT). Algorithms for MaxSAT are still fairly inefficient in industrial settings, in part because the most effective SAT techniques cannot be easily extended to MaxSAT. This paper proposes a novel algorithm for MaxSAT that improves existing state of the art solvers by orders of magnitude on industrial benchmarks. The new algorithm exploits modern SAT solvers, being based on the identification of unsatisfiable subformulas. Moreover, the new algorithm provides additional insights between unsatisfiable subformulas and the maximum satisfiability problem.

The results of the last MaxSAT Evaluations suggest there is no universal best algorithm for solving MaxSAT, as the fastest solver often depends on the type of instance. Having an oracle able to predict the most suitable MaxSAT solver for a given instance would result in the most robust solver. Inspired by the success of SATzilla for SAT, this paper describes the first approach for a portfolio of algorithms for MaxSAT. Compared to existing solvers, the resulting portfolio can achieve significant performance improvements on a representative set of instances.

Exact Max-SAT solvers, compared with SAT solvers, apply little inference at each node of the proof tree. Commonly used SAT inference rules like unit propagation produce a simplified formula that preserves satisfiability but, unfortunately, solving the Max-SAT problem for the simplified formula is not equivalent to solving it for the original formula. In this paper, we define a number of original inference rules that, besides being applied efficiently, transform Max-SAT instances into equivalent Max-SAT instances which are easier to solve. The soundness of the rules, that can be seen as refinements of unit resolution adapted to Max-SAT, are proved in a novel and simple way via an integer programming transformation. With the aim of finding out how powerful the inference rules are in practice, we have developed a new Max-SAT solver, called MaxSatz, which incorporates those rules, and performed an experimental investigation. The results provide empirical evidence that MaxSatz is very competitive, at least, on random Max-2SAT, random Max-3SAT, Max-Cut, and Graph 3-coloring instances, as well as on the benchmarks from the Max-SAT Evaluation 2006.