Search

A report from the University of Oslo discusses the use of machine learning algorithms and heuristic stopping criteria in systematic reviews in education and educational psychology. The study found that using machine learning algorithms for abstract screening reduced the workload of irrelevant records by an average of 58% and saved an estimated 1.66 days of time. The random forests algorithm with sentence bidirectional encoder representations from transformers performed the best. The study also found that heuristic stopping rules could retrieve 95% of relevant abstracts, with the best results achieved by stopping the screening process after classifying 20% of records and consecutively classifying 5% of irrelevant papers. The research provides empirical evidence on the performance of machine learning screening algorithms in this field. [Extracted from the article]

A study conducted by researchers at Beijing Jiaotong University in China has developed a new algorithm called EnergySaving Maximum Entropy Deep Reinforcement Learning (ES-MEDRL) to optimize the energy consumption of urban rail transit systems. The algorithm incorporates Lagrange multipliers and maximum policy entropy to minimize train traction energy consumption while balancing ride comfort, punctuality, and safety. Numerical experiments conducted on the Beijing Metro's Yizhuang Line showed a minimum 20% increase in energy efficiency compared to manual driving. The algorithm has the potential to guide energy-efficient train operations at the planning level. [Extracted from the article]

A new report from Shaanxi Normal University in Xi'an, China, presents research on an effective metaheuristic algorithm called the Inherited Optimization Algorithm (IOA). Inspired by the cognitive tendencies and adaptive feedback behavior of people, the IOA algorithm aims to solve complex optimization problems in engineering applications. The algorithm's performance was evaluated on various benchmarks and compared to other metaheuristics, demonstrating its computational efficiency and search efficiency. The research was funded by the National Natural Science Foundation of China and the Fundamental Research Funds for the Central Universities. [Extracted from the article]

A recent study conducted at Xi'an University of Science and Technology in Xi'an, People's Republic of China, focused on the accurate recognition of micro-expressions in demanding scenarios such as clinical psychotherapy and criminal interrogation. The study proposed a novel architecture based on a multi-scale 3D residual convolutional neural network, which effectively captured weak and fleeting facial features and improved recognition performance. The algorithm's performance was evaluated using public datasets, and it achieved recognition accuracies of 74.6%, 84.77%, and 91.35% on these datasets. The study concluded that the proposed algorithm could serve as an important reference for researchers working on high-precision micro-expression recognition. [Extracted from the article]

A recent study conducted by researchers at Sichuan University in Chengdu, China explores the use of computer algorithms to mimic the route choices made by human players in wargames. The researchers developed a fuzzy dynamic route choice model and algorithm that takes into account the uncertain and dynamic decision-making environment of wargaming. The model was applied to a practical case and demonstrated its effectiveness in handling fuzzy and uncertain information while simulating the decision-making psychology of wargame players. This research provides a new method for automatic route choice in wargame simulations. [Extracted from the article]

A study conducted by researchers at Kunsan National University in South Korea explores the construction of a positive psychological fitness assessment model for athletes. The study utilizes a system engineering method and DM method to analyze and study athletes' positive psychological fitness. The model, based on Fuzzy C-clustering (FCM) and Secure Neural Network, demonstrates improved accuracy in psychological crisis analysis compared to other algorithms. The findings suggest that the model has the potential for self-learning and online self-diagnosis of athletes' positive psychological disorders, providing valuable support to psychological counseling and fitness teams in universities. The study also highlights the importance of utilizing big data for studying new approaches to pre-alarm psychological fitness crises. [Extracted from the article]

A recent report from Indiana University discusses the use of wisdom of the crowd algorithms to label medical images. The study tested various algorithms that relied on the input of individuals recruited through an app-based platform. The best-performing algorithms were found to select top performers, weight decisions by training accuracy, and consider the task environment. These algorithms outperformed expert performance and have implications for the development of medical artificial intelligence. The full article can be accessed for free through the provided link. [Extracted from the article]

A recent study conducted at Ural Federal University in Ekaterinburg, Russia, explores the optimization of energy management in renewable-based multi-microgrid systems. The study proposes a modified optimization algorithm called Modified Student Psychology-Based Optimization (MSPBO) to address issues such as slow convergence and low solution accuracy. The MSPBO algorithm is applied to a two-stage multi-objective optimization problem, aiming to minimize operating costs and maximize operator benefits while integrating renewable sources and hydrogen energy storage systems. The research concludes that the integration of hydrogen energy storage systems leads to a notable enhancement in peak load reduction. [Extracted from the article]

A research study conducted at the University of Catania in Italy explores a new approach to Interactive Evolutionary Multiobjective Optimization (IEMO) in the field of Psychology and Psychiatry. The study proposes a method called eXplainable Interactive Multiobjective Evolutionary optimization Approach (XIMEA), which combines preference elicitation and optimization phases to find the best compromise solution for conflicting objectives. The Decision Maker (DM) is involved in the process and their preferences are translated into guidelines for the algorithm. The research concludes that XIMEA-DRSA is effective in converging towards the most interesting part of the Pareto front. [Extracted from the article]