Matthew Martinez

Gottfried Wilhelm Leibniz is the self-proclaimed inventor of the binary system and is considered as such by most historians of mathematics and/or mathematicians. Really though, we owe the groundwork of today’s computing not to Leibniz but to the Englishman Thomas Harriot and the Spaniard Juan Caramuel de Lobkowitz, whom Leibniz plagiarized. This plagiarism has been identified on the basis of several facts: Caramuel’s work on the binary system is earlier than Leibniz’s, Leibniz was acquainted—both directly and indirectly—with Caramuel’s work and Leibniz had a natural tendency to plagiarize scientific works.

Very few scientists today question the fact that information, together with matter and energy, is one of the three constructs forming the ontology of the universe. However, there is still a long way to go before in order to establish the interrelations between information and energy and information and matter, as Einstein did between matter and energy. In this paper, after introducing the energy, matter, information model, which covers the three constructs and their relationships, we illustrate real examples—two qualitative and two quantitative—of the interrelations between energy and information. This closes the open question with respect to the interrelationship between energy and information.

A partir de una encuesta por correo electrónico a estudiantes de Ingeniería en el SAI de la UAMA en un contexto de enseñanza remota forzado por la pandemia, se caracteriza la gestión de la docencia, desde la visión de los estudiantes, se realiza un diagnóstico de los aspectos relacionados. Es un estudio transversal con un análisis cualitativo y descriptivo de las respuestas. La gestión docente se ve afectada por factores inherentes al proceso, así como también por factores externos. Los resultados obtenidos muestran que fue impactada fuertemente al realizarse el proceso totalmente vía remota especialmente, la comunicación e interacción alumno-profesor.

Previous research has focused on the relation between social class and prosocial behavior. However, this relation is yet unclear. In this work, we shed light on this issue by considering the effect of the level of empathy and the social class of the recipient of help on two types of prosociality, namely helping and caring. In one experimental study, we found that for high-class participants, empathy had a positive effect on helping, regardless of the recipient’s social class. However, empathy had no effect for low-class participants. When it comes to caring, empathy had a positive effect for both high and low-class participants, but only when the recipient of help belonged to the same social class. This highlights that empathy by itself is not sufficient to promote cooperative relations and that the social class of the recipient of help should be taken into account to shed light on this issue.

Feature selection is a widely studied technique whose goal is to reduce the dimensionality of the problem by removing irrelevant features. It has multiple benefits, such as improved efficacy, efficiency and interpretability of almost any type of machine learning model. Feature selection techniques may be divided into three main categories, depending on the process used to remove the features known as Filter, Wrapper and Embedded. Embedded methods are usually the preferred feature selection method that efficiently obtains a selection of the most relevant features of the model. However, not all models support an embedded feature selection that forces the use of a different method, reducing the efficiency and reliability of the selection. Neural networks are an example of a model that does not support embedded feature selection. As neural networks have shown to provide remarkable results in multiple scenarios such as classification and regression, sometimes in an ensemble with a model that includes an embedded feature selection, we attempt to embed a feature selection process with a general-purpose methodology. In this work, we propose a novel general-purpose layer for neural networks that removes the influence of irrelevant features. The Feature-Aware Drop Layer is included at the top of the neural network and trained during the backpropagation process without any additional parameters. Our methodology is tested with 17 datasets for classification and regression tasks, including data from different fields such as Health, Economic and Environment, among others. The results show remarkable improvements compared to three different feature selection approaches, with reliable, efficient and effective results.

Time series forecasting is a well-known deep learning application field in which previous data are used to predict the future behavior of the series. Recently, several deep learning approaches have been proposed in which several nonlinear functions are applied to the input to obtain the output. In this paper, we introduce a novel method to improve the performance of deep learning models in time series forecasting. This method divides the model into hierarchies or levels from simpler to more complex ones. Simpler levels handle smoothed versions of the input, whereas the most complex level processes the original time series. This method follows the human learning process where general/simpler tasks are performed first, and afterward, more precise/harder ones are accomplished. Our proposed methodology has been applied to the LSTM architecture, showing remarkable performance in various time series. In addition, a comparison is reported including a standard LSTM and novel methods such as DeepAR, Temporal Fusion Transformer, NBEATS and Echo State Network.