Abstract Quantum Machine Learning (QML) holds the promise of improving conventional machine learning, but the conditions under which an advantage can be obtained are still unclear. Although many studies have been conducted in the literature, there is no conclusive evidence that allows us to determine which types of classical datasets or which problem complexities benefit from the use of quantum techniques. Moreover, there have been contradicting findings when dealing with small or unbalanced datasets. In order to systematically approach this challenge and clarify the need for quantum properties in specific tasks, we propose the adoption of a framework capable of exploring the solution space of quantum feature maps and providing solid evidence of their potential advantages. Our framework comprises three main components: (a) the construction of datasets designed to highlight the specific characteristics that are expected to be relevant in quantum machine learning tasks; (b) the definition of metrics that serve as proxies for quantum kernels with desirable properties; and (c) an evolutionary-guided search for quantum feature maps maximizing those metrics. When using this framework with sensible parameter choices, we obtain results suggesting that some previous studies may have reported overfitted outcomes. This shows that justifying the need for quantum mechanical properties might be beyond the actual scope of conventional classical tasks, since there is no clear quantum feature that contributes to the gain shown by some of the QML techniques most commonly applied in the literature.

Abstract Quantum machine learning, positioned at the convergence of quantum computing and artificial intelligence, is an emerging and highly promising field, primarily due to its potential to enhance the performance of classical machine learning systems. As this area is developing at an exceptionally rapid pace, it is essential to remain up to date with the latest advancements and research. This paper introduces pLazyQML, a software package designed to accelerate, automate, and streamline experimentation with quantum machine learning models on classical computers. pLazyQML reduces the complexity and time required for developing and testing quantum-enhanced machine learning models. Comprehensive experiments on established models and datasets demonstrate the efficiency, scalability, and workflow simplification provided by pLazyQML, making it a valuable tool for researchers and practitioners in quantum machine learning.

Abstract Sound source separation (SSS) is a fundamental problem in audio signal processing, aiming to recover individual audio sources from a given mixture. A promising approach is multichannel non-negative matrix factorization (MNMF), which employs a Gaussian probabilistic model encoding both magnitude correlations and phase differences between channels through spatial covariance matrices (SCM). In this work, we present a dedicated hardware architecture implemented on field programmable gate arrays (FPGAs) for efficient SSS using MNMF-based techniques. A novel decorrelation constraint is presented to facilitate the factorization of the SCM signal model, tailored to the challenges of multichannel source separation. The performance of this FPGA-based approach is comprehensively evaluated, taking advantage of the flexibility and computational capabilities of FPGAs to create an efficient real-time source separation framework. Our experimental results demonstrate consistent, high-quality results in terms of sound separation.

Abstract The accurate estimation of respiratory rate (RR) is crucial for assessing the respiratory system’s health in humans, particularly during auscultation processes. Despite the numerous automated RR estimation approaches proposed in the literature, challenges persist in accurately estimating RR in noisy environments, typical of real-life situations. This becomes especially critical when periodic noise patterns interfere with the target signal. In this study, we present a parallel driver designed to address the challenges of RR estimation in real-world environments, combining multi-core architectures with parallel and high-performance techniques. The proposed system employs a nonnegative matrix factorization (NMF) approach to mitigate the impact of noise interference in the input signal. This NMF approach is guided by pre-trained bases of respiratory sounds and incorporates an orthogonal constraint to enhance accuracy. The proposed solution is tailored for real-time processing on low-power hardware. Experimental results across various scenarios demonstrate promising outcomes in terms of accuracy and computational efficiency.

Abstract Classical music, unlike popular music, is usually recorded live with close microphone techniques. For this reason, isolated tracks are not available to create the final mixture/stream, and so the mixing process requires greater effort. Source separation methods are a potential solution to this problem. However, current algorithms are not fast enough to yield real-time separation in professional setups with dozens of microphones and sources. In this paper, we propose a fast approach consisting of a panning-based multichannel non-negative matrix factorization model to separate classical music. We tested the system on real professional recordings, where we were able to reach real-time with very low latency and promising quality.

Abstract In this paper, a parallel computing method is proposed to perform the background denoising and wheezing detection from a multi-channel recording captured during the auscultation process. The proposed system is based on a non-negative matrix factorization (NMF) approach and a detection strategy. Moreover, the initialization of the proposed model is based on singular value decomposition to avoid dependence on the initial values of the NMF parameters. Additionally, novel update rules to simultaneously address the multichannel denoising while preserving an orthogonal constraint to maximize source separation have been designed. The proposed system has been evaluated for the task of wheezing detection showing a significant improvement over state-of-the-art algorithms when noisy sound sources are present. Moreover, parallel and high-performance techniques have been used to speedup the execution of the proposed system, showing that it is possible to achieve fast execution times, which enables its implementation in real-world scenarios.

High Performance Computing Clusters (HPCCs) are essential platforms for solving up-to-date challenges through parallel and distributed applications. Nevertheless, HPCCs have an important economic and environmental impact owing to the large amounts of energy required for their operation. In this work, an improved version of our EECluster software focused on reducing the operating costs and emissions of HPCCs is presented. Multicriteria learning algoritms are leveraged to jointly optimize cluster performance, direct and indirect operating costs both from an economic and an environmental standpoint. EECluster is a mature software that can be integrated with the resource management systems OGE/SGE and PBS/TORQUE.

Over the last years, research has been placed at the core of numerous software products resulting from the ubiquitous presence of multimedia devices in our lives. Besides the need of acquiring specialized hardware, performing reliable and reproducible research requires the exploitation of specialized tools that are not traditionally present in the agile enterprise ecosystem. This might be a factor that ultimately draws many organizations away from introducing research into their software products. In this work, we propose a hands-free workflow that exploits widely available tools in the enterprise, like source control (SC) and continuous integration (CI) systems. We demonstrate that this workflow acts as a single-step end-to-end solution, maximizing the usage of the available hardware, and ensuring the repeatability of the performed experiments. The probability of human errors is minimized by automating all file transfers, and feedback is provided at the end of every trial with the location of the results. Generated artifacts are automatically archived, alongside the initial conditions of the experiment allowing for its full recreation. Although in our solution we exploit Git and Jenkins, this workflow can also be implemented with any of the SC and CI tools typically available in the enterprise.

A fuzzy rule-based classifier is proposed in this paper where the number of rules in the knowledge base that are fired when an object is classified is anti-monotone with respect to the prior probability of its class. This classifier is intended to secure an equilibrium between accuracy and energy consumption, which is critical in battery operated embedded devices. The method is compared to legacy multi-criteria evolutionary algorithms, where a group of classifiers with different balances between accuracy and consumption are evolved, and the most accurate classifier is selected among those individuals in the Pareto front whose use of the battery does not exceed a given threshold. A significant increase in the battery life is reported without a degradation in the quality of service.

Deep neural networks (DNNs) have achieved unprecedented success in a wide array of tasks. However, the performance of these systems depends directly on their hyper-parameters which often must be selected by an expert. Optimizing the hyper-parameters remains a substantial obstacle in designing DNNs in practice. In this work, we propose to select them using particle swarm optimization (PSO). Such biologically-inspired approaches have not been extensively exploited for this task. We demonstrate that PSO efficiently explores the solution space, allowing DNNs of a minimal topology to obtain competitive classification performance over the MNIST dataset. We showed that very small DNNs optimized by PSO retrieve promising classification accuracy for CIFAR-10. Also, PSO improves the performance of existing architectures. Extensive experimental study, backed-up with the statistical tests, revealed that PSO is an effective technique for automating hyper-parameter selection and efficiently exploits computational resources.

Many real-world classification systems must comply with a series of inherent restrictions to the problem at hand such as response times, power consumptions or computational costs. This poses a fundamental limitation to traditional performance-driven classifiers and learning algorithms by restraining their applicability in cost-sensitive scenarios. Because of this, fuzzy systems are leveraged to learn cost-conscious multi-stage classifiers through multiobjective optimization to find a set of optimal tradeoffs between accuracy and any related cost. This approach allows find a suitable balance between all objectives regardless of the scenario. Experimental evaluations were done for Sound Environment Classification in modern battery-powered hearing aids by jointly optimising classification accuracy and computational costs.

Today, High Performance Computing clusters (HPC) are an essential tool owing to they are an excellent platform for solving a wide range of problems through parallel and distributed applications. Nonetheless, HPC clusters consume large amounts of energy, which combined with notably increasing electricity prices are having an important economical impact, forcing owners to reduce operation costs. In this work we propose a software, named EECluster, to reduce the high energy consumption of HPC clusters. EECluster works with both OGE/SGE and PBS/TORQUE resource management systems and automatically tunes its decision-making mechanism based on a machine learning approach. The quality of the obtained results using this software are evaluated by means of experiments made using actual workloads from the Scientific Modelling Cluster at Oviedo University and the academic-cluster used by the Oviedo University for teaching high performance computing subjects.

Feature Selection is an important task within Text Categorization, where irrelevant or noisy features are usually present, causing a lost in the performance of the classifiers. Feature Selection in Text Categorization has usually been performed using a filtering approach based on selecting the features with highest score according to certain measures. Measures of this kind come from the Information Retrieval, Information Theory and Machine Learning fields. However, wrapper approaches are known to perform better in Feature Selection than filtering approaches, although they are time-consuming and sometimes infeasible, especially in text domains. However a wrapper that explores a reduced number of feature subsets and that uses a fast method as evaluation function could overcome these difficulties. The wrapper presented in this paper satisfies these properties. Since exploring a reduced number of subsets could result in less promising subsets, a hybrid approach, that combines the wrapper method and some scoring measures, allows to explore more promising feature subsets. A comparison among some scoring measures, the wrapper method and the hybrid approach is performed. The results reveal that the hybrid approach outperforms both the wrapper approach and the scoring measures, particularly for corpora whose features are less scattered over the categories.

Text categorization, which consists of automatically assigning documents to a set of categories, usually involves the management of a huge number of features. Most of them are irrelevant and others introduce noise which could mislead the classifiers. Thus, feature reduction is often performed in order to increase the efficiency and effectiveness of the classification. In this paper, we propose to select relevant features by means of a family of linear filtering measures which are simpler than the usual measures applied for this purpose. We carry out experiments over two different corpora and find that the proposed measures perform better than the existing ones.

Abstract Text Categorization is the process of assigning documents to a set of previously fixed categories. A lot of research is going on with the goal of automating this time‐consuming task. Several different algorithms have been applied, and Support Vector Machines (SVM) have shown very good results. In this report, we try to prove that a previous filtering of the words used by SVM in the classification can improve the overall performance. This hypothesis is systematically tested with three different measures of word relevance, on two different corpus (one of them considered in three different splits), and with both local and global vocabularies. The results show that filtering significantly improves the recall of the method, and that also has the effect of significantly improving the overall performance.