The subset selection problem is fundamental in machine learning and other fields of computer science. We introduce a stochastic formulation for the minimum cost subset selection problem in a black box setting, in which only the subset metric value is available. Subsequently, we can handle two-stage schemes, with an outer subset-selection component and an inner subset cost evaluation component. We propose formulating the subset selection problem in a stochastic manner by choosing subsets at random from a distribution whose parameters are learned. Two stochastic formulations are proposed. The first explicitly restricts the subset's cardinality, and the second yields the desired cardinality in expectation. The distribution is parameterized by a decision variable, which we optimize using Stochastic Mirror Descent. Our choice of distributions yields constructive closed-form unbiased stochastic gradient formulas and convergence guarantees, including a rate with favorable dependency on the problem parameters. Empirical evaluation of selecting a subset of layers in transfer learning complements our theoretical findings and demonstrates the potential benefits of our approach.

Web application security has become a major concern in recent years, as more and more content and services are available online. A useful method for identifying security vulnerabilities is black-box testing, which relies on an automated crawling of web applications. However, crawling Rich Internet Applications (RIAs) is a very challenging task. One of the key obstacles crawlers face is the state similarity problem: how to determine if two client-side states are equivalent. As current methods do not completely solve this problem, a successful scan of many real-world RIAs is still not possible. We present a novel approach to detect redundant content for security testing purposes. The algorithm applies locality-sensitive hashing using MinHash sketches in order to analyze the Document Object Model (DOM) structure of web pages, and to efficiently estimate similarity between them. Our experimental results show that this approach allows a successful scan of RIAs that cannot be crawled otherwise.

Motivation: The complexes formed by binding of proteins to RNAs play key roles in many biological processes, such as splicing, gene expression regulation, translation and viral replication. Understanding protein-RNA binding may thus provide important insights to the functionality and dynamics of many cellular processes. This has sparked substantial interest in exploring protein-RNA binding experimentally, and predicting it computationally. The key computational challenge is to efficiently and accurately infer protein-RNA binding models that will enable prediction of novel protein-RNA interactions to additional transcripts of interest. Results: We developed DLPRB (Deep Learning for Protein-RNA Binding), a new deep neural network (DNN) approach for learning intrinsic protein-RNA binding preferences and predicting novel interactions. We present two different network architectures: a convolutional neural network (CNN), and a recurrent neural network (RNN). The novelty of our network hinges upon two key aspects: (i) the joint analysis of both RNA sequence and structure, which is represented as a probability vector of different RNA structural contexts; (ii) novel features in the architecture of the networks, such as the application of RNNs to RNA-binding prediction, and the combination of hundreds of variable-length filters in the CNN. Our results in inferring accurate RNA-binding models from high-throughput in vitro data exhibit substantial improvements, compared to all previous approaches for protein-RNA binding prediction (both DNN and non-DNN based). A more modest, yet statistically significant, improvement is achieved for in vivo binding prediction. When incorporating experimentally-measured RNA structure, compared to predicted one, the improvement on in vivo data increases. By visualizing the binding specificities, we can gain biological insights underlying the mechanism of protein RNA-binding. Availability and implementation: The source code is publicly available at https://github.com/ilanbb/dlprb. Supplementary information: Supplementary data are available at Bioinformatics online.

MOTIVATION: New sequencing technologies generate larger amount of short reads data at decreasing cost. De novo sequence assembly is the problem of combining these reads back to the original genome sequence, without relying on a reference genome. This presents algorithmic and computational challenges, especially for long and repetitive genome sequences. Most existing approaches to the assembly problem operate in the framework of de Bruijn graphs. Yet, a number of recent works use the paradigm of string graph, using a variety of methods for storing and processing suffixes and prefixes, like suffix arrays, the Burrows-Wheeler transform or the FM index. Our work is motivated by a search for new approaches to constructing the string graph, using alternative yet simple data structures and algorithmic concepts. RESULTS: We introduce a novel hash-based method for constructing the string graph. We use incremental hashing, and specifically a modification of the Karp-Rabin fingerprint, and Bloom filters. Using these probabilistic methods might create false-positive and false-negative edges during the algorithm's execution, but these are all detected and corrected. The advantages of the proposed approach over existing methods are its simplicity and the incorporation of established probabilistic techniques in the context of de novo genome sequencing. Our preliminary implementation is favorably comparable with the first string graph construction of Simpson and Durbin (2010) (but not with subsequent improvements). Further research and optimizations will hopefully enable the algorithm to be incorporated, with noticeable performance improvement, in state-of-the-art string graph-based assemblers.