526 publications from this institution
Genetic studies of Mendelian and rare diseases face the critical challenges of identifying pathogenic gene variants and their modes-of-action. Previous efforts rarely utilized the tissue-selective manifestation of these diseases for their elucidation. Here we introduce an interpretable machine learning (ML) platform that utilizes heterogeneous and large-scale tissue-aware datasets of human genes, and rigorously, concurrently and quantitatively assesses hundreds of candidate mechanisms per disease. The resulting tissue-aware ML platform is applicable in gene-specific, tissue-specific, or patient-specific modes. Application of the platform to selected Mendelian disease genes pinpointed mechanisms that lead to tissue-specific disease manifestation. When applied jointly to diseases that manifest in the same tissue, the models revealed common known and previously underappreciated factors that underlie tissue-selective disease manifestation. Lastly, we harnessed our ML platform toward genetic diagnosis of tissue-selective rare diseases. Patient-specific models of candidate disease-causing genes from 50 patients successfully prioritized the pathogenic gene in 86% of the cases, implying that the tissue-selectivity of rare diseases aids in filtering out unlikely candidate genes. Thus, interpretable tissue-aware ML models can boost mechanistic understanding and genetic diagnosis of tissue-selective heritable diseases. A webserver supporting gene prioritization is available at https://netbio.bgu.ac.il/trace/ .
No abstract is provided for this article.
Recommender systems (RS) are a class of information filter applications whose main goal is to provide personalized recommendations, content, and services to users. Recommendation services may support a firm's marketing strategy and contribute to increase revenues. Most RS methods were designed to provide recommendations of single items. Generating bundle recommendations, i.e., recommendations of two or more items together, can satisfy consumer needs, while at the same time increase customers’ buying scope and the firm's income. Thus, finding and recommending an optimal and personal bundle becomes very important. Recommendation of bundles of products should also involve personalized pricing to predict which price should be offered to a user in order for the bundle to maximize purchase probability. However, most recommendation methods do not involve such personal price adjustment. This paper introduces a novel model of bundle recommendations that integrates collaborative filtering (CF) techniques, demand functions, and price modeling. This model maximizes the expected revenue of a recommendation list by finding pairs of products and pricing them in a way that maximizes both the probability of its purchase by the user and the revenue received by selling the bundle. Experiments with several real-world datasets have been conducted in order to evaluate the accuracy of the bundling model predictions. This paper compares the proposed method with several state-of-the-art methods (collaborative filtering and SVD). It has been found that using bundle recommendation can improve the accuracy of results. Furthermore, the suggested price recommendation model provides a good estimate of the actual price paid by the user and at the same time can increase the firm's income.
No abstract is provided for this article.
In this paper, we propose a new unsupervised approach for identifying suspicious access to sensitive relational data. In the proposed method, a tree-like model encapsulates the characteristics of the result-set (i.e., data) that the user normally access within each possible context. During the detection phase, result-sets are examined against the induced model and a similarity score is derived.
No abstract available.
Deep neural networks are being used to solve complex classification problems, in which other machine learning classifiers, such as SVM, fall short. Recurrent Neural Networks (RNNs) have been used for tasks that involves sequential inputs, like speech to text. In the cyber security domain, RNNs based on API calls have been able to classify unsigned malware better than other classifiers. In this paper we present a black-box attack against RNNs, focusing on finding adversarial API call sequences that would be misclassified by a RNN without affecting the malware functionality. We also show that the this attack is effective against many classifiers, due-to the transferability principle between RNN variants, feed-forward DNNs and state-of-the-art traditional machine learning classifiers. Finally, we introduce the transferability by transitivity principle, causing an attack against generalized classifier like RNN variants to be transferable to less generalized classifiers like feed-forward DNNs. We conclude by discussing possible defense mechanisms.
Improving the robustness of neural nets in regression tasks is key to their application in multiple domains. Deep learning-based approaches aim to achieve this goal either by improving their prediction of specific values (i.e., point prediction), or by producing prediction intervals (PIs) that quantify uncertainty. We present IPIV, a deep neural network for producing both a PI and a value prediction. Our loss function expresses the value prediction as a function of the upper and lower bounds, thus ensuring that it falls within the interval without increasing model complexity. Moreover, our approach makes no assumptions regarding data distribution within the PI, making its value prediction more effective for various real-world problems. Experiments and ablation tests on known benchmarks show that our approach produces tighter uncertainty bounds than the current state-of-the-art approaches for producing PIs, while maintaining comparable performance to the state-of-the-art approach for value-prediction. Additionally, we go beyond previous work and include large image datasets in our evaluation, where IPIV is combined with modern neural nets.
Decision forests are known to excel in tabular data once their hyperparameters are well-tuned. In addition to being accurate and robust classifiers, these models can be easily converted into a collection of if-else rules that clearly describe the model’s decision-making process. However, in practice, decision trees may reach enormous depth and size, and as a result, the collection of rules is vast and complicated. Furthermore, the model may consume a significant amount of memory space on the machine. By creating compact models, it is possible to establish a modest set of rules that may be easier to understand, require less memory, and by nature, may increase the decision-making ability and avoid overfitting. Previous studies attempted to reduce the size of the trees by altering their structure, but this affects both their advantages and simplicity because their structure is much more complicated. (e.g. oblique trees). In this research, we present FACET, a novel algorithm for retaining the compactness of trees while referring to the model as a black box. FACET addresses this by utilizing automated feature engineering methods, which generate a new feature set from the data set by manipulating the current feature set, resulting in a drastic reduction in the size of the decision trees while preserving and even improving the model’s accuracy. Our algorithm, FACET, has been extensively tested on multiple datasets, models, and operators to demonstrate its effectiveness. On average, FACET achieves a 24% reduction in the size criteria of the tree-based model without sacrificing accuracy. This reduction in size leads to an average memory reduction of 44% on the dataset required for learning. These statistically significant results demonstrate the potential of FACET to enable more efficient and interpretable tree-based models, without compromising their accuracy, in practical applications.
Transfer Learning aims to transfer knowledge from a source task to a target task. We focus on a situation when there is a large number of available source models, and we are interested in choosing a single source model that can maximize the predictive performance in the target domain. Existing methods compute some form of "similarity" between the source task data and the target task data. They then select the most similar source task and use the model trained on it for transfer learning. Previous methods do not account for the fact that it is the model parameters that are transferred rather than the data. Therefore, the "similarity" of the source data does not directly influence transfer learning performance. In addition, we would like the possibility of confidently selecting a source model even when the data it was trained on is not available, for example, due to privacy or copyright constraints. We propose to use the truncated source models as encoders for the target data. We then select a source model based on how well it clusters the target data in the latent encoding space, which we calculate using the Mean Silhouette Coefficient. We prove that if the encodings achieve a Mean Silhouette Coefficient of 1, optimal classification can be achieved using just the final layer of the target network. We evaluate our method using the University of California, Riverside (UCR) time series archive and show that the proposed method achieves comparable results to previous work, without using the source data.
No abstract is provided for this article.
No abstract is provided for this article.
The data clustering problem can be described as the task of organizing data into groups, where in each group the objects share some similar attributes. Most of the problems clustering algorithms address do not have a prior solution. This paper addresses the algorithm selection challenge for data clustering, while taking the difficulty in evaluating clustering solutions into account. We present a new meta-learning method for recommending the most suitable clustering algorithm for a dataset. Based on concepts from the isolation forest algorithm, we propose a new similarity measure between datasets. Our proposed dataset characterization methods generate an embedding for a dataset using this similarity measure, which is then used to improve the quality of the problem’s characterization. The method utilizes landmarking concepts to characterize the dataset and then, inspired by the DeepFM algorithm, applies meta-learning to rank the candidate algorithms that are expected to perform the best for the current dataset. This ranking could, among other things, support AutoML systems. Our approach is evaluated on a corpus of 100 publicly available benchmark datasets. We compare our method’s ranking performance to that of existing meta-learning methods and show the dominance of our method in terms of predictive performance and computational complexity.
