10 June 2021 in IEEE Latin America Transactions
The COVID-19 is a new disease from the virus SARS-CoV-2, the infection can cause respiratory illness with symptoms such as cough, fever, and, in severe cases, pneumonia. Early diagnosis is crucial for the correct treatment to reduce as much as possible the stress in the healthcare system. The need for auxiliary diagnostic tools has increased as there are no accurate automated toolkits available. Application of advanced artificial intelligence (AI) techniques coupled with radiological imaging can be helpful for the accurate detection of this disease. In this study, we have applied learning transfer to a convolutional neural network known as AlexNet for binary chest X-ray recognition (COVID-19 vs Healthy). We have fine-tunned AlexNet for our specific problem. The first layer, which works with RGB images, is replaced for images in a single intensity (grayscale). 11,312 chest X-ray images from six public databases were used to train the network. Among them are samples of healthy people and samples that present the effect of pneumonia and COVID-19 diseases. The results prove that deep learning with chest X-ray images can extract significant biomarkers related to COVID-19, since the obtained accuracy, sensitivity and specificity were 96.5%, 98.0%, and 91.7%, respectively. ROC analysis and confusion matrices are used to validate the results of the fine-tunned AlexNet network
15 August 2019 in Energy Science & Engineering
JDJosé Rafael Dorrego‐Portela
José Rafael Dorrego‐Portela
MAMaximino Avendaño‐Alejo
+1
The development of an optical contact instrument for measuring the geometric shape of aerodynamic profiles on blades of small power wind turbines is presented. The instrument uses the triangulation principle, where a structured line laser pattern is projected onto the surface of one of the faces of the blade under test, a camera captures the image of the line and is processed to interpret the distorted form of the projected line. A linear sweep of the instrument makes it possible to measure the profile in another section of the blade. Comparison and evaluation results of two symmetric profiles of the NACA 0012 family are presented, one manufactured in a 3D printer and the other one is a metal profile AF104 of a subsonic wind tunnel. Additionally, three sections of a blade with profile FX 63‐137 of a 1.5 kW wind turbine were evaluated. An aerodynamic analysis shows a reduction in the lift coefficient and in the efficiency of the aerodynamic profile, as well as an increase in the drag coefficient. The sensitivity of the instrument is 0.1 mm on the Z‐axis.
12 May 2022 in European Journal of Physics
There are many physical systems (with randomness) where the performance of the probability distribution is of great importance [for example the so-called radioactive dice (Murray and Hart 2012 Phys. Educ. 47 197–201)]. A relevant situation is when the value (positive integers) of the upper faces is added when rolling fair dice (unbiased), that is, balanced, regular or unshaven. It has been widely studied in the basic literature for one, two and a maximum of four dice. However, a generalization using these methods for many dice is too cumbersome to carry out. In this paper we explore a method that uses the moment generating function, already studied in Singh et al (2011 UNLV Gaming Res. Rev. J.), for five six-sided fair dice with positive integer values. We carry out the generalization of this method to study the performance of the probability distribution and make a quick counting without complications for the sum of n fair dice with s sides (s ⩾ 6), and real values. We also study the invariance of the distribution when we set the number of sides, that is, for s-fix. Therefore, we study the distributions for the outgoing total sum (random variable) of roll generalized dice, for combinations of n multi-sided dice.
9 July 2025 in European Journal of Physics
In a previous paper (2022 Eur. J. Phys. 43 045801), we carried out a study on quickly calculating the probability distributions of many dice with a number of faces greater than 6 (it can also be less). This formulation allows us to unambiguously obtain the graphical and tabular distribution of the random variable of interest—that is, X = ∑
j
X
j
—where we were interested in the total sum of the resulting sides of the dice. We also involve the concept of radioactive dice, giving a more physical and didactic approach. We generalize the moment–generating function (MGF) method to include real and imaginary numbers. Now, in this complementary work we research the situation in which an unconventional dice system has a grouped or clustered bias; that is, arbitrarily choosing the probabilistic weight assigned to each side of the dice, which can happen absolutely randomly or be manipulated. We show the mathematical characteristics—statistical–probabilistic—that must be fulfilled, as well as examples with simulations for different cases, which illustrate the formulation effectiveness. We introduce the concept of cluster bias and the cluster bias probability distribution function (additive cluster bias and multiplicative inverse cluster bias). Also, we give a brief perspective of applications in other scientific branches.
1 October 2026 in European Journal of Physics
Entropy is one of the most fundamental concepts in physics and information theory (Shannon 1948 Bell Syst. Tech. J. 27 623–56), its correct understanding is essential for the study of various physical systems, especially in thermodynamics, statistical mechanics, and in thermal process engineering. This blending of concepts from physics and information theory has made it difficult for many students, even specialists, to clearly grasp the power, scope, and proper domain of the applicability of the theory, particularly at the undergraduate and graduate levels, mainly due to the abundance of highly artificial theoretical models. In this work, we present a systematic and comprehensive study of the entropy applicability, starting from elementary random systems such as fair (unbiased) and biased dice throws. Specifically, we examine a single dice with an arbitrary number of faces, i.e. . We build upon previous studies (Sánchez-Sánchez et al 2022 Eur. J. Phys. 43 045801; Sánchez-Sánchez and Cortés-p 2025 Eur. J. Phys. 46 045704) that serve as a basis to develop a mathematically consistent analysis, while avoiding ambiguities and unnecessary formal excesses. We compute the corresponding entropy measures for generalized cases involving biased distributions using the probability functions derived in these previous works—additionally the Escher-type distribution for demonstrating its equivalence with the Gibbs–Boltzmann distribution — from which the unbiased cases naturally arise as particular limits. Thus, we provide more significant insight into both the analytical development and the visual interpretation of the resulting entropy metrics. This theoretical framework can be used to introduce advanced undergraduate or postgraduate students to the fundamental characteristics of biased random systems (distributions) and their intrinsic entropy, which are more general than standard cases.