Article
Materials Science, Multidisciplinary
Ravindranadh Bobbili
Summary: The study suggests using machine learning models to predict the glass forming ability of BMGs, and compares the accuracies of different models. The results show that the XGB model is more efficient than the ANN model, with higher accuracy and precision. The study also finds that characteristic temperature plays a significant role in understanding the glass formation of alloys.
Article
Clinical Neurology
Mohamed Sobhi Jabal, Olivier Joly, David Kallmes, George Harston, Alejandro Rabinstein, Thien Huynh, Waleed Brinjikji
Summary: This study developed machine learning models using clinical and imaging features to predict the functional outcome at 3 months after thrombectomy in acute ischemic stroke patients. Combining clinical and imaging features resulted in the best prediction. Age, NIHSS score, degree of brain atrophy, early ischemic core, and collateral circulation deficit volume on CTA were the most important classifying features.
FRONTIERS IN NEUROLOGY
(2022)
Article
Critical Care Medicine
Emma Schwager, Erina Ghosh, Larry Eshelman, Kalyan S. Pasupathy, Erin F. Barreto, Kianoush Kashani
Summary: We developed and validated two parsimonious algorithms to predict the time of diagnosis of any stage of acute kidney injury (any-AKI) or moderate-to-severe AKI in clinically actionable prediction windows. The models achieved good discriminative performance and high positive predictive values in predicting AKI risk at least 6 hours (any-AKI) and 12 hours (moderate-to-severe AKI) prior to diagnosis in ICU patients.
JOURNAL OF CRITICAL CARE
(2023)
Review
Biochemistry & Molecular Biology
Xiaoyin Li, Xiao Liu, Xiaoyan Deng, Yubo Fan
Summary: Cardiovascular disease (CVD) is the most common cause of morbidity and mortality worldwide. Recent advancements in artificial intelligence (AI), especially machine learning (ML), have enabled the prediction of CVD. This review summarizes the applications of ML in cardiovascular diseases, including direct prediction of CVD based on risk factors or medical imaging findings, and indirect assessment of CVD using ML-based hemodynamics with vascular geometries. It also discusses the use of ML models as surrogates for computational fluid dynamics, accelerating the disease prediction process and reducing manual intervention.
Article
Biochemical Research Methods
Yongxian Fan, Xiqian Lu, Guicong Sun
Summary: This study aims to predict whether a patient has hepatitis C using different machine learning models and provides explanations for the prediction process. The proposed method significantly outperforms existing methods in terms of predictive performance and maintains excellent model interpretability, enabling clinicians to better understand the model's decision-making process.
BMC BIOINFORMATICS
(2023)
Article
Computer Science, Artificial Intelligence
Stella Dimitsaki, George I. Gavriilidis, Vlasios K. Dimitriadis, Pantelis Natsiavas
Summary: This article evaluates a set of machine learning algorithms for predicting the severity of COVID-19 patients based on plasma proteomics and clinical data. The use of an ensemble of ML algorithms is designed and deployed for early patient triage. The evaluation shows that MLP and SVM algorithms perform the best.
ARTIFICIAL INTELLIGENCE IN MEDICINE
(2023)
Article
Computer Science, Information Systems
Javier Enrique Camacho-Cogollo, Isis Bonet, Bladimir Gil, Ernesto Iadanza
Summary: Sepsis is a deadly syndrome with heterogeneous clinical manifestation. Early diagnosis and appropriate treatment are crucial for improving patient survival. Although several prediction models have been proposed, their efficacy is limited. Therefore, developing models for sepsis prediction using machine learning techniques is of great importance.
Article
Medicine, General & Internal
Joung Ouk (Ryan) Kim, Yong-Suk Jeong, Jin Ho Kim, Jong-Weon Lee, Dougho Park, Hyoung-Seop Kim
Summary: This study developed a cardiovascular diseases prediction model using machine learning algorithms based on health screening datasets, with gradient boosting, extreme gradient boosting, and random forest algorithms showing the best performance among all tested. Machine learning algorithms improved CVD prediction accuracy compared to previous models.
Article
Computer Science, Artificial Intelligence
Mara Graziani, Lidia Dutkiewicz, Davide Calvaresi, Jose Pereira Amorim, Katerina Yordanova, Mor Vered, Rahul Nair, Pedro Henriques Abreu, Tobias Blanke, Valeria Pulignano, John O. Prior, Lode Lauwaert, Wessel Reijers, Adrien Depeursinge, Vincent Andrearczyk, Henning Muller
Summary: Since its emergence in the 1960s, Artificial Intelligence (AI) has been widely applied to various technology products and fields. Machine learning, as a major part of current AI solutions, achieves high performance on various tasks through learning from data and experience. However, the interpretability of AI models, especially deep neural networks, is often challenging. Different domains have different requirements for interpretability and tools for debugging and validating models. In this paper, the authors propose a unified terminology and definition of interpretability in AI systems, aiming to improve clarity and efficiency in the regulation of ethical and reliable AI development, and to facilitate communication across interdisciplinary areas of AI.
ARTIFICIAL INTELLIGENCE REVIEW
(2023)
Article
Materials Science, Multidisciplinary
Sohaib Nazar, Jian Yang, Muhammad Nasir Amin, Kaffayatullah Khan, Muhammad Ashraf, Fahid Aslam, Mohammad Faisal Javed, Sayed M. Eldin
Summary: This study developed empirical models for predicting compressive strength (CS) and slump values of fly ash-based geopolymer concrete using three artificial intelligence-based algorithms - adaptive neuro-fuzzy inference system (ANFIS), artificial neural networks (ANNs), and gene expression programming (GEP). The GEP model outperformed the ANFIS and ANN models in terms of R-value, R2, and RMSE. The GEP model generated more accurate predictions for slump and CS after rigorous training and optimization of hyperparameters.
JOURNAL OF MATERIALS RESEARCH AND TECHNOLOGY-JMR&T
(2023)
Article
Computer Science, Artificial Intelligence
Supreeth P. Shashikumar, Christopher S. Josef, Ashish Sharma, Shamim Nemati
Summary: Sepsis is a major cause of morbidity and mortality in ICU, and early prediction using DeepAISE, an artificial intelligence model, has shown high accuracy and low false alarm rates compared to other baseline models in hospitalized patients.
ARTIFICIAL INTELLIGENCE IN MEDICINE
(2021)
Review
Cardiac & Cardiovascular Systems
Evangelos K. Oikonomou, Rohan Khera
Summary: Artificial intelligence and machine learning have the potential to revolutionize healthcare, particularly in the management of diabetes and its cardiovascular complications. This review provides an overview of the various data-driven methods and their application in personalized care for diabetes patients at increased cardiovascular risk. The article discusses the role of artificial intelligence in diagnosis, prognostication, phenotyping, and treatment, as well as the challenges and ethical considerations that arise. It also emphasizes the need for regulatory standards to ensure the effectiveness and safety of medical artificial intelligence products in transforming cardiovascular care and outcomes in diabetes.
CARDIOVASCULAR DIABETOLOGY
(2023)
Article
Engineering, Industrial
Antonio L. L. Alfeo, Mario G. C. A. Cimino, Gigliola Vaglini
Summary: Predictive maintenance (PdM) utilizes machine learning technologies to monitor asset health and plan maintenance activities. However, some health-related measures may not be sufficient to assess health stage reliably. This study addresses this issue by combining a health stage classifier with a feature learning mechanism to extract informative features from minimally processed data.
JOURNAL OF MANUFACTURING SYSTEMS
(2022)
Review
Computer Science, Interdisciplinary Applications
Aizatul Shafiqah Mohd Faizal, T. Malathi Thevarajah, Sook Mei Khor, Siow-Wee Chang
Summary: Cardiovascular disease is a global health issue, with AI approach replacing traditional statistical models for risk prediction. Biomarkers aid in early detection and risk assessment, but current research faces various problems and challenges.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2021)
Article
Business, Finance
Dangxing Chen, Jiahui Ye, Weicheng Ye
Summary: Forecasting credit default risk is an important research field. Logistic regression has been widely recognized as a solution due to its accuracy and interpretability. However, we introduce a neural network with a selective option to increase interpretability. Our methods are tested on two datasets and we find that, for most samples, logistic regression is sufficient, but a shallow neural network model provides better accuracy without significantly sacrificing interpretability for specific data portions.
RESEARCH IN INTERNATIONAL BUSINESS AND FINANCE
(2023)
Article
Computer Science, Artificial Intelligence
Daniel Gonzalez, Miguel A. Patricio, Antonio Berlanga, Jose M. Molina
Summary: According to the World Health Organization, the global population over 60 is expected to double by 2050, with a majority preferring to live alone. It is important to find mechanisms and tools, such as smart home systems, to help improve their autonomy.
Article
Computer Science, Information Systems
Jesus Iriz, Miguel A. Patricio, Antonio Berlanga, Jose M. Molina
Summary: This paper presents a neural model capable of equalizing songs according to the musical genre and adapting to changes and nuances in the music.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Miguel A. Patricio, Antonio Berlanga, David Palomero, Jose M. Molina
Summary: One of the problems in online marketing in omni-channel environments is efficient budget allocation. A new multi-criteria attribution model (MAMOM) is proposed to address this issue, which takes into account expert opinions and experiences from different departments, resulting in more accurate investment allocation.
INTERNATIONAL JOURNAL OF INFORMATION TECHNOLOGY & DECISION MAKING
(2022)
Article
Engineering, Electrical & Electronic
Mohamad Ghaddar, Jose-Maria Molina-Garcia-Pardo, Ismail Ben Mabrouk, Martine Lienard, Pierre Degauque
Summary: Limited previous work has been done on deterministic modeling of 5G millimeter-wave wireless communications in nonuniform tunnels. This study predicts and analyzes broadband MIMO propagation in underground mine tunnels for 5G and beyond. The tunnel surfaces are modeled as diffracting rectangular prism wedges, and a deterministic RT MIMO model based on UTD is developed. The predicted MIMO subchannels are successfully fitted with experimental measurements, and the MIMO multipath fading in underground tunnels is found to be lognormally distributed and dependent on tunnel size and shape.
IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION
(2023)
Article
Chemistry, Analytical
Ricardo Robles-Enciso, Isabel Pilar Morales-Aragon, Alfredo Serna-Sabater, Maria Teresa Martinez-Ingles, Antonio Mateo-Aroca, Jose-Maria Molina-Garcia-Pardo, Leandro Juan-Llacer
Summary: In this research, power and quality measurements of four transmissions using different emission technologies were conducted in an indoor environment at the frequency of 868 MHz under two non-line-of-sight (NLOS) conditions. The received power of a narrowband continuous wave (CW) signal, as well as the Received Signal Strength Indicator (RSSI) and bit error rate (BER) of LoRa and Zigbee signals, were measured. Quality parameters of a 20 MHz bandwidth 5G QPSK signal were also measured. Two fitting models, the Close-in (CI) model and the Floating-Intercept (FI) model, were used to analyze the path loss. The results revealed different slopes for the NLOS-1 and NLOS-2 zones, and the FI model achieved the best accuracy in both NLOS situations.
Article
Computer Science, Information Systems
Jose-Victor Rodriguez, Maria-Teresa Martinez-Ingles, Jose-Maria Molina Garcia-Pardo, Leandro Juan-Llacer, Ignacio Rodriguez-Rodriguez
Summary: This paper presents two uniform theories of diffraction-physical optics (UTD-PO) formulations to analyze radiowave multiple diffraction in vegetated urban areas with both buildings and trees. The solutions consider buildings modeled as knife-edges and rectangular sections, and the effect of tree canopy is taken into account by adding proper attenuation factors/phasors. The validation of these formulations has been done through comparisons and measurements on a scaled-model. The advantage of the solutions is that they only include single diffractions, avoiding higher-order diffraction terms and reducing computational requirements.
Article
Computer Science, Information Systems
Alvaro Luis Bustamante, Miguel A. Patricio, Antonio Berlanga, Jose M. Molina
Summary: This article presents an ML workflow based on ML operations (MLOps) over the Thinger.io IoT platform to streamline the transition from model training to model deployment on edge devices. Edge computing extends cloud computing capabilities by bringing services near the edge of a network, supporting a new variety of AI services and ML applications.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Computer Science, Artificial Intelligence
Asier Alcaide, Miguel A. Patricio, Antonio Berlanga, Angel Arroyo, Juan J. Cuadrado Gallego
Summary: Facial verification has made breakthroughs in recent years, but its application expansion is limited by the complex computing requirements of deep learning models. This paper proposes a new lightweight model that can solve the problem of running facial verification on low computing resource devices.
INTERNATIONAL JOURNAL OF INTERACTIVE MULTIMEDIA AND ARTIFICIAL INTELLIGENCE
(2022)
Proceedings Paper
Engineering, Electrical & Electronic
Leandro Juan-Llacer, Jose Maria Molina-Garcia-Pardo, Alain Sibille, Saul A. Torrico, Luis Martinez Rubiola, Maria Teresa Martinez-Ingles, Jose-Victor Rodriguez, Juan Pascual-Garcia
Summary: This study analyzes the path loss of wireless communication systems in agriculture using different frequency bands. The results indicate the need to consider the multiple-scattering contributions from trees and the observed guiding effect.
2022 16TH EUROPEAN CONFERENCE ON ANTENNAS AND PROPAGATION (EUCAP)
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Daniel Amigo, David Sanchez Pedroche, Jesus Garcia, Jose M. Molina
Summary: Geographic Information Systems (GIS) allow analysis based on geo-referenced data. This study aims to develop an automatic framework for extracting geo-referenced trees by merging different data sources and has achieved satisfactory results.
16TH INTERNATIONAL CONFERENCE ON SOFT COMPUTING MODELS IN INDUSTRIAL AND ENVIRONMENTAL APPLICATIONS (SOCO 2021)
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Juan Pedro Llerena, Jesus Garcia, Jose Manuel Molina
Summary: Ship type identification in maritime context is crucial for authorities to control activities. This paper proposes a deep learning approach for binary classification to identify fishing ships, showing better performance compared to traditional machine learning methods and balancing techniques.
16TH INTERNATIONAL CONFERENCE ON SOFT COMPUTING MODELS IN INDUSTRIAL AND ENVIRONMENTAL APPLICATIONS (SOCO 2021)
(2022)
