Abstract With the latest advances in deep learning-based generative models, it has not taken long to take advantage of their remarkable performance area time series. Deep neural networks used work with series heavily depend on size and consistency datasets training. These features are usually abundant real world, where they limited often have constraints that must be guaranteed. Therefore, an effective way increase amount data is by using augmentation techniques, either adding noise or...

Abstract Harmful algal blooms (HABs) are a growing concern to public health and aquatic ecosystems. Long-term water monitoring conducted by hand poses several limitations the proper implementation of safety plans. This work combines automatic high-frequency (AFHM) systems with machine learning (ML) techniques build data-driven chlorophyll- (Chl- ) soft-sensor. Massive data for temperature, pH, electrical conductivity (EC) system battery were taken three years at intervals 15 min from two...

Providing useful information to the users by recommending highly demanded products and services is a fundamental part of business many top tier companies. Recommender Systems make use sources provide with accurate predictions novel recommendations items. Here we propose, DeepMF, collaborative filtering method that combines Deep Learning paradigm Matrix Factorization (MF) improve quality both made user. Specifically, DeepMF performs successive refinements MF model layered architecture uses...

The lack of bias management in Recommender Systems leads to minority groups receiving unfair recommendations.Moreover, the trade-off between equity and precision makes it difficult obtain recommendations that meet both criteria.Here we propose a Deep Learning based Collaborative Filtering algorithm provides with an optimum balance fairness accuracy.Furthermore, recommendation stage, this does not require initial knowledge users' demographic information.The proposed architecture incorporates...

Abstract IoT edge computing is a new paradigm “in the domain” for performing calculations and processing at of network, closer to user source data. This relatively recent, and, together with cloud fog computing, there may be some confusion about its meaning implications. paper aims help practitioners researchers better understand what industry thinks is, expected benefits challenges associated this paradigm. We conducted survey using semi-structured in-depth questionnaire collect qualitative...

Abstract Due to the growing rise of cyber attacks in Internet, demand accurate intrusion detection systems (IDS) prevent these vulnerabilities is increasing. To this aim, Machine Learning (ML) components have been proposed as an efficient and effective solution. However, its applicability scope limited by two important issues: (i) shortage network traffic data datasets for attack analysis, (ii) privacy constraints be used. overcome problems, Generative Adversarial Networks (GANs) synthetic...

<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Context</i> : DevOps can be defined as a cultural movement to improve and accelerate the delivery of business value by making collaboration between development operations effective. xmlns:xlink="http://www.w3.org/1999/xlink">Objective</i> This paper aims help practitioners researchers better understand organizational structure characteristics teams adopting DevOps....

Neural collaborative filtering is the state of art field in recommender systems area; it provides some models that obtain accurate predictions and recommendations. These are regression-based, they just return rating predictions. This paper proposes use a classification-based approach, returning both their reliabilities. The extra information (prediction reliabilities) can be used variety relevant areas such as detection shilling attacks, recommendations explanation or navigational tools to...

The research on empirical software engineering that uses qualitative data analysis is increasing. However, most of them do not deepen into the validity findings, specifically in reliability coding which these methodologies rely on. This paper aims to establish a novel theoretical framework enables methodological approach for conducting this through Inter-Coder Agreement (ICA), based use coefficients measure degree agreement collaborative coding. We systematically review several existing...

The qualitative research on empirical software engineering that uses Grounded Theory is increasing (GT). trustworthiness, rigor, and transparency of GT data analysis can benefit, among others, when multiple analysts juxtapose diverse perspectives collaborate to develop a common code frame based consensual consistent interpretation. Inter-Rater Reliability (IRR) and/or Agreement (IRA) are commonly used techniques measure consensus, thus shared However, minimal guidance available about how...

Solving the convergence issues of Generative Adversarial Networks (GANs) is one most outstanding problems in generative models. In this work, we propose a novel activation function to be used as output generator agent. This based on Smirnov probabilistic transformation and it specifically designed improve quality generated data. sharp contrast with previous works, our provides more general approach that deals not only replication categorical variables but any type data distribution...

DevOps is becoming a main competency required by the software industry. However, academic institutions have been slow to provide training in engineering (SE) curricula. One reason for this fact that problems addressed may be hard understand students who not previously worked industry or on projects of meaningful size and complexity. This paper shows an experience integrates SE curricula through research-based teaching (RBT). We aim expose led companies adopt researching analyzing real cases...

Recommender systems aim to estimate the judgment or opinion that a user might offer an item. Matrix-factorization-based collaborative filtering typifies both users and items as vectors of factors inferred from item rating patterns. This method finds latent structure in data, assuming observations lie close low-dimensional space. However, matrix factorizations have been traditionally designed by hand. Here, we present Evolutionary Matrix Factorization (EMF), evolutionary approach...

Gender-based crime is one of the most concerning scourges contemporary society, and governments worldwide have invested lots economic human resources to foretell their occurrence anticipate aggressions. In this work, we propose apply Machine Learning (ML) techniques create models that accurately predict recidivism risk a gender-violence offender. We feed model with data extracted from official Spanish VioGen system comprising more than 40,000 reports gender violence. To evaluate performance,...

Abstract Deep learning provides accurate collaborative filtering models to improve recommender system results. matrix factorization and their related neural networks are the state of art in field; nevertheless, both lack necessary stochasticity create robust, continuous, structured latent spaces that variational autoencoders exhibit. On other hand, data augmentation through autoencoder does not provide results field due high sparsity systems. Our proposed apply concept inject space deep...

Traditionally, recommender systems have been approached as regression models aiming to predict the score that a user would give particular item. In this work, we propose system tackles problem classification task instead of regression. The new model, Dirichlet Matrix Factorization (DirMF), provides not only prediction but also its reliability, hence achieving better balance between quality and quantity predictions (i.e., reducing error by limiting model’s coverage). experimental results...

This PhD Thesis is devoted to the study of Hodge structures on a special type complex algebraic varieties, so-called character varieties. For this purpose, we propose use powerful algebro-geometric tool coming from theoretical physics, known as Topological Quantum Field Theory (TQFT). With idea in mind, present develop formalism that allows us construct TQFTs two simpler pieces data: field theory (geometric data) and quantisation (algebraic data). As an application, TQFT computing...

In this paper, we use lax monoidal TQFTs as an effective computational method for motivic classes of representation varieties. particular, perform the calculation parabolic $\mathrm{SL}_2(\mathbb{C})$-representation varieties over a closed orientable surface arbitrary genus and any number marked points with holonomies Jordan type. This technique is based on building physical inspiration that generalizes construction González-Prieto, Logares Muñoz.