Article
Chemistry, Multidisciplinary
Ghada Elkhawaga, Mervat Abu-Elkheir, Manfred Reichert
Summary: Predictive process monitoring (PPM) is an important application of process mining that utilizes machine learning to predict the future of ongoing business processes. However, the need for explainable artificial intelligence (XAI) to gain user trust in the predictions remains a challenge. This study systematically investigates the effects of different choices in PPM settings on the explainability of generated predictions, highlighting inconsistencies between data characteristics, ML model predictors, and prediction explanations.
APPLIED SCIENCES-BASEL
(2022)
Article
Engineering, Environmental
Jinxi Dong, Zhaosheng Yu, Xikui Zhang, Jiajun Luo, Qihong Zou, Chao Feng, Xiaoqian Ma
Summary: Under the pressure of energy and environmental protection, new energy vehicles are becoming the future direction of automotive development. The safety performance of the power battery has always been the most critical indicator in the new energy vehicle industry. A model that can predict the battery life can be obtained using Machine Learning. This paper discusses the importance assessment of features, the hyperparameter search process, and the comparison of different algorithms, and concludes that CatBoost has the highest prediction accuracy.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2023)
Article
Engineering, Mechanical
Marko Percic, Sasa Zelenika, Igor Mezic
Summary: A recent study systematically characterized nanoscale friction in technological thin films using machine learning algorithms, improving prediction accuracy and providing an effective tool for further scientific and technological nanotribological analyses.
Article
Robotics
Kong Yao Chee, Tom Z. Jiahao, M. Ani Hsieh
Summary: This letter presents a method to enhance the dynamic models used in model predictive control (MPC) for quadrotor control using deep learning. By integrating a first-principle model and a neural network, the hybrid model can accurately predict the quadrotor dynamics and demonstrates improved performance in closed-loop control.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2022)
Article
Biochemical Research Methods
Ye Wang, Tathagata Bhattacharya, Yuchao Jiang, Xiao Qin, Yue Wang, Yunlong Liu, Andrew J. Saykin, Li Chen
Summary: With the development and decreasing cost of next-generation sequencing technologies, the study of the human microbiome has become a rapidly expanding research field with various clinical applications. Building a prediction model for clinical outcomes based on microbiome data is essential for improving prediction performance. The phylogenetic tree represents a unique correlation structure of microbiome and can be important for this purpose.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Chemistry, Multidisciplinary
David Orrego Granados, Jonathan Ugalde, Rodrigo Salas, Romina Torres, Javier Linkolk Lopez-Gonzales
Summary: The academic success of university students depends on multiple factors, and a novel visual-predictive data analysis approach was used in this study to provide insights into students' academic performance and aid in policy development.
APPLIED SCIENCES-BASEL
(2022)
Review
Biotechnology & Applied Microbiology
Walker D. D. Short, Oluyinka O. O. Olutoye, Benjamin W. W. Padon, Umang M. M. Parikh, Daniel Colchado, Hima Vangapandu, Shayan Shams, Taiyun Chi, Jangwook P. P. Jung, Swathi Balaji
Summary: Impaired wound healing is a significant financial and medical burden. Disruptions in ECM deposition, structure, and composition lead to impaired healing in diseased states, such as in diabetes. Valid measures include lack of bacterial contamination, good tissue perfusion, and reduced mechanical injury and strain. This review discusses bioengineering advances in detecting biologic and physiologic factors, visualizing and modeling the ECM, and efficiently evaluating wound data. It focuses on bioelectronics and biologic interfaces for real-time sensing and actuation, advanced imaging techniques, and computational modeling. These advancements have the potential to improve healing outcomes and guide decision-making in wound care.
FRONTIERS IN BIOENGINEERING AND BIOTECHNOLOGY
(2022)
Article
Engineering, Multidisciplinary
Aleksandar Pejic, Piroska Stanic Molcer
Summary: This study investigates the potential of deriving a predictive model for a problem-solving process from raw log-files, using data from OECD's PISA 2012 computer-based assessment database. Two feature sets were extracted from the dataset and evaluated with machine learning algorithms, aiming to understand problem-solving patterns and improve e-learning systems for training such skills.
ACTA POLYTECHNICA HUNGARICA
(2021)
Article
Chemistry, Multidisciplinary
Chuang Liu, Haojie Wang, Yingkui Du, Zhonghu Yuan
Summary: Student achievement prediction is a crucial research direction in educational data mining, which directly reflects students' mastery of courses and teachers' teaching level. This paper proposes a student achievement prediction model based on evolutionary spiking neural network and demonstrates its accuracy through experimental results.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Artificial Intelligence
Qinbin Li, Zeyi Wen, Zhaomin Wu, Sixu Hu, Naibo Wang, Yuan Li, Xu Liu, Bingsheng He
Summary: With increasing concern about data privacy, federated learning has become a popular research topic for collaborative training of machine learning models under privacy restrictions. This survey provides a comprehensive review of federated learning systems, introducing key system components and analyzing their design. It also presents a categorization of federated learning systems based on six aspects, including data distribution, machine learning model, privacy mechanism, communication architecture, scale of federation, and motivation. The categorization can guide the design of federated learning systems, as demonstrated by the case studies. The survey summarizes existing federated learning systems and offers insights into design factors, case studies, and future research opportunities.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Agronomy
Joao Vasco Silva, Joost van Heerwaarden, Pytrik Reidsma, Alice G. Laborte, Kindie Tesfaye, Martin K. van Ittersum
Summary: The performance of statistical and machine learning methods in explaining and predicting crop yield variability was assessed in this study. The results showed that big data from farmers' fields can to some extent explain on-farm yield variability, but not predict it across time and space.
FIELD CROPS RESEARCH
(2023)
Article
Physics, Multidisciplinary
Mehmet Salti, Emel Ciger, Evrim Ersin Kangal, Bilgin Zengin
Summary: In this study, we redesigned the generalized pressure dark energy model using a caloric framework and employed machine learning techniques to analyze the cosmic Hubble parameter. The optimized model parameters were obtained using a genetic neural network algorithm and the most recent observational measurements. Additionally, we addressed the issue of calculating errors on the optimized parameter values using the Fisher Information Matrix algorithm. The results showed good agreement with observational data and provided additional cosmological insights.
Review
Computer Science, Information Systems
Dalia Abdulkareem Shafiq, Mohsen Marjani, Riyaz Ahamed Ariyaluran Habeeb, David Asirvatham
Summary: Student retention is a crucial metric in education, and various techniques such as educational data mining and learning analytics are employed to improve teaching practices and identify at-risk students. However, there are challenges in applying predictive models and incorporating important factors like heterogeneous and homogeneous student groups.
Article
Computer Science, Theory & Methods
Martin Husak, Vaclav Bartos, Pavol Sokol, Andrej Gajdos
Summary: This paper discusses predictive methods in cyber defense and evaluates three distinct approaches in a common environment. The methods, including data mining, dynamic network entity reputation scoring, and time series analysis, have shown promising accuracy and usability for predicting and projecting ongoing cyberattacks.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2021)
Review
Computer Science, Artificial Intelligence
Alican Dogan, Derya Birant
Summary: This paper provides a comprehensive analysis on applying machine learning techniques in manufacturing, addressing unanswered research questions and summarizing key methods and algorithms used for improving manufacturing processes over the past two decades. Additionally, it discusses in detail the main technological advances and applications in manufacturing, as well as providing insights into the current state of the industry.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)