Examinando por Autor "Lancho, Carmen"

Mostrando 1 - 8 de 8

A complexity measure for binary classification problems based on lost points
(Springer International Publishing, 2021) Lancho, Carmen; Martín de Diego, Isaac; Cuesta, Marina; Aceña, Víctor; M. Moguerza, Javier
Complexity measures are focused on exploring and capturing the complexity of a data set. In this paper, the Lost points (LP) complexity measure is proposed. It is obtained by applying k-means in a recursive and hierarchical way and it provides both the data set and the instance perspective. On the instance level, the LP measure gives a probability value for each point informing about the dominance of its class in its neighborhood. On the data set level, it estimates the proportion of lost points, referring to those points that are expected to be misclassified since they lie in areas where its class is not dominant. The proposed measure shows easily interpretable results competitive with measures from state-of-art. In addition, it provides probabilistic information useful to highlight the boundary decision on classification problems.
From classification to visualization: a two way trip
(Springer International Publishing, 2021) Cuesta, Marina; Martín de Diego, Isaac; Lancho, Carmen; Aceña, Víctor; M. Moguerza, Javier
High Dimensional Data (HDD) is one of the biggest challenges in Data Science arising from Big Data. The application of dimensionality reduction techniques over HDD allows visualization and, thus, a better problem understanding. In addition, these techniques also can enhance the performance of Machine Learning (ML) algorithms while increasing the explanatory power. This paper presents an automatic method capable of obtaining an adequate representation of the data, given a previously trained ML model. Likewise, an automatic method is introduced to bring a Support Vector Machine (SVM) model based on an adequate representation of the data. Both methods provide an Explanaible Machine Learning procedure. The proposal is tested on several data sets providing promising results. It significantly eases the visualization and understanding task to the data scientist when a ML model has already been trained, as well as the ML selection parameters when a reduced representation of data has been achieved.
Health Sufficiency Indicators for Pandemic Monitoring
(MDPI, 2021) M. Moguerza, Javier; Perelló Oliver, Salvador; Martín de Diego, Isaac; Aceña, Víctor; Lancho, Carmen; Cuesta, Marina; González Fernández, César
The outbreak of the COVID-19 disease, spreading all around the world and causing a worldwide pandemic, has lead to the collapse of the health systems of the most affected countries. Due to the ease of transmission, early prevention measures are proved to be fundamental to control the pandemic and, hence, the saturation of the health systems. Given the difficulty of obtaining characteristics of these systems of different countries and regions, it is necessary to define indicators based on basic information that enable the assessment of the evolution of the impact of a disease in a health system along with fair comparisons among different ones. This present paper introduces the Health Sufficiency Indicator (HSI), in its accumulated and daily versions. This indicator measures the additional pressure that a health care system has to deal with due to a pandemic. Hence, it allows to evaluate the capacity of a health system to give response to the corresponding needs arising from a pandemic and to compare the evolution of the disease among different regions. In addition, the Potential Occupancy Ratio (POR) in both its hospital ward bed and ICU bed versions is here introduced to asses the impact of the pandemic in the capacity of hospitals. These indicators and other well-known ones are applied to track the evolution of the impact of the disease on the Spanish health system during the first wave of the pandemic, both on national and regional levels. An international comparison among the most affected countries is also performed.
Hostility measure for multi-level study of data complexity
(Springer, 2022) Lancho, Carmen; Martín De Diego, Isaac; Cuesta, Marina; Aceña, Víctor; Moguerza, Javier M.
Complexity measures aim to characterize the underlying complexity of supervised data. These measures tackle factors hindering the performance of Machine Learning (ML) classifiers like overlap, density, linearity, etc. The state-of-the-art has mainly focused on the dataset perspective of complexity, i.e., offering an estimation of the complexity of the whole dataset. Recently, the instance perspective has also been addressed. In this paper, the hostility measure, a complexity measure offering a multi-level (instance, class, and dataset) perspective of data complexity is proposed. The proposal is built by estimating the novel notion of hostility: the difficulty of correctly classifying a point, a class, or a whole dataset given their corresponding neighborhoods. The proposed measure is estimated at the instance level by applying the k-means algorithm in a recursive and hierarchical way, which allows to analyze how points from different classes are naturally grouped together across partitions. The instance information is aggregated to provide complexity knowledge at the class and the dataset levels. The validity of the proposal is evaluated through a variety of experiments dealing with the three perspectives and the corresponding comparative with the state-of-the-art measures. Throughout the experiments, the hostility measure has shown promising results and to be competitive, stable, and robust.
Padel two-dimensional tracking extraction from monocular video recordings
(Springer, 2024-11-14) Novillo, Álvaro; Aceña, Víctor; Lancho, Carmen; Cuesta, Marina; Martín de Diego, Isaac
This study introduces a novel framework for the automatic two-dimensional tracking of padel games using monocular recordings. By integrating advanced Computer Vision and Deep Learning techniques, our algorithm detects and tracks players, the court, and the ball. Through homography, we accurately project detected player positions onto a twodimensional court, enabling comprehensive tracking throughout the game. We tested the proposed algorithm using amateur video recordings of padel games found in literature. This approach remains user-friendly, cost-effective, and adaptable to various camera angles and lighting conditions. This makes it accessible to both amateur and professional players and coaches, providing a valuable tool for performance analysis. Additionally, the proposed framework holds potential for adaptation to other sports with minimal modifications, further broadening its applicability.
Real-time classification of cattle behavior using Wireless Sensor Networks
(Elsevier, 2024) Navarro, Jorge; R. Fernández, Rubén; Aceña, Víctor; Fernández-Isabel, Alberto; Lancho, Carmen; Martín de Diego, Isaac
La detección de patrones de actividad y comportamiento utilizando acelerómetros en humanos ha sido una línea de investigación prolongada. Los avances en este campo se han transferido con éxito al estudio del comportamiento animal gracias a la aparición de nuevas tecnologías del Internet de las Cosas (IoT), como las Redes de Sensores Inalámbricos (WSN), y a la necesidad de información comportamental más compleja. Todos los sistemas propuestos por la comunidad científica han sido evaluados en términos de rendimiento de clasificación. Sin embargo, no muchos estudios consideran la posible pérdida de precisión cuando estos sistemas se implementan en WSN, dada la baja capacidad computacional de sus nodos y la necesidad de un bajo consumo energético. Este artículo propone un sistema de clasificación de patrones de comportamiento para cuatro tipos de comportamiento animal en ganado de pastoreo libre, junto con una configuración óptima y una configuración restringida del mismo. La evaluación de este sistema tiene en cuenta su rendimiento de clasificación y su precisión esperada bajo los recursos limitados que pueden ofrecer las WSN. Los resultados muestran que la configuración óptima mejora el rendimiento de sus alternativas en un promedio del 9% y la configuración restringida en un promedio del 6%. Además, como parte de una WSN, los resultados demuestran una precisión impecable en las configuraciones óptima y restringida para caminar (100% y 100%), casi perfecta para pastar (98.39% y 98.59%), y aceptable para acostarse (79.03% y 69.01%) y estar de pie (75.81% y 70.42%). En conclusión, el sistema propuesto representa una herramienta poderosa para analizar comportamientos complejos en el ganado mediante el uso de WSN.
Selección de Ejercicios de Lógica: para ciberseguridad e inteligencia artificial (2a Edición)
(2023-09-06) Arias, Joaquín; Lancho, Carmen; Ramírez, Iván
Selección de Ejercicios de Lógica: para ciberseguridad e inteligencia artificial.
(2022) Arias, Joaquín; Lancho, Carmen
Selección de ejercicios de lógica proposicional y de primer orden.