Earticle

With the growing interest in both public and private 5G services based on millimeter wave (mmWave) communication for indoor usage scenarios such as smart factories, site design specialists are seeking sophisticated methods and tools for simulating indoor radio coverage based on highly accurate path loss prediction models. Although machine learning approaches can be used in path loss modeling thanks to the highly accurate prediction capability, their performance can be limited by the size of available measurement data set used for training. In this paper, we propose new approaches to train path loss models in the few-shot learning scenarios of smart factories. The proposed approaches are based on meta-learning with slight modifications to perform fine-tuning over an entire train data set rather than a meta-test data set. It is shown that the indoor path loss models based on convolutional neural networks (CNNs) trained by meta-learning based on three different meta-train task assignment schemes outperform both a conventional CNN model and an empirical model.

An online social network is a platform that is continuously expanding, which enables groups of people to share their views and communicate with one another using the Internet. The social relations among members of the public are significantly improved because of this gesture. Despite these advantages and opportunities, criminals are continuing to broaden their attempts to exploit people by making use of techniques and approaches designed to undermine and exploit their victims for criminal activities. The field of digital forensics, on the other hand, has made significant progress in reducing the impact of this risk. Even though most of these digital forensic investigation techniques are carried out manually, most of these methods are not usually appropriate for use with online social networks due to their complexity, growth in data volumes, and technical issues that are present in these environments. In both civil and criminal cases, including sexual harassment, intellectual property theft, cyberstalking, online terrorism, and cyberbullying, forensic investigations on social media platforms have become more crucial. This study explores the use of machine learning techniques for addressing criminal incidents on social media platforms, particularly during forensic investigations. In addition, it outlines some of the difficulties encountered by forensic investigators while investigating crimes on social networking sites.

Artificial intelligence, especially deep learning technology, is penetrating the majority of research areas, including the field of bioinformatics. However, deep learning has some limitations, such as the complexity of parameter tuning, architecture design, and so forth. In this study, we analyze these issues and challenges in regards to its applications in bioinformatics, particularly genomic analysis and medical image analytics, and give the corresponding approaches and solutions. Although these solutions are mostly rule of thumb, they can effectively handle the issues connected to training learning machines. As such, we explore the tendency of deep learning technology by examining several directions, such as automation, scalability, individuality, mobility, integration, and intelligence warehousing.

Invasive species are organisms that are introduced into places outside of their natural distribution range. The global pet trade is facilitating the introduction of invasive species into new countries and areas. Among the introduced alien species, turtles are one of the most common animal groups whether lives in wetland ecosystems, such as wetlands or reservoirs. Like other countries around the world, exotic turtles is becoming a growing concern for the wetland ecosystem in South Korea. In this study, we report new reports of subspecies of Painted turtle (Chrysemys spp.): Chrysemys picta marginata, C. p. bellii and C. dorsalis, from the reservoirs in downtown Cheongju and Gwangju, South Korea. We used morphological features, such as the characteristics of the legs, plastron, and carapace, to identify the turtles. It is assumed that all turtles were artificially released into nature. Considering the increasing number of reports on the introduction of alien invasive turtles in Korean wetlands, we recommend the formulation of an immediate and systematic management plan for pet trades and organized continuous monitoring programs.

This study proposes a machine learning approach to understand how post-earnings-announcement drift (PEAD) works. We analyze when PEAD, combined with other factors, becomes more pronounced. To accommodate diverse variables and more complex specifications, two tree-based machine learning approaches including eXtreme Gradient Boosting (XGBoost) and Light Gradient Boosting Machine (LightGBM) are used to examine the relationship between PEAD and 89 variables. The long-short portfolio produced by LightGBM model reports 2.1 times higher returns than the portfolio’s returns, based on the conventional measure of earnings surprise. The model enhances the economic and statistical significance of the long-short portfolio returns. SHapley Additive exPlanations (SHAP) analysis determines feature importance and shows that liquidity, firm size, profitability ratios, share turnover, net trading flows by retail investors, and earnings surprises, play an important role in the prediction of PEAD.

In an effort to raise funds, North Korea now performs hacking assaults against the world's financial institutions. More specifically, the North Korean hackers demand money to decrypt the files they created, and since these transactions are handled anonymously, it is difficult to identify them. Therefore, this research uses the BitcoinHeist dataset to identify cryptocurrency-related ransomware. We construct the experiment with two distinct steps: classification and anomaly detection. The XG boosting technique achieved a 100% accuracy score in the first experiment. Even though anomaly detection methods were used in the second trial for detection, they only managed to get a precision score of 50%, whereas XG boosting produced 92%. These tests indicate that the machine learning method for ransomware detection is effective. This study excels in classification and anomaly detection, which is especially noteworthy given that another paper recently classified ransomware variants except for the "white" designation.

For this study a new memory-based vocabulary strategic learning framework was constructed involving four essential stages for vocabulary learning which can be translated into four corresponding strategy-driven processes that execute learning actions. A questionnaire study was conducted with Chinese university EFL learners who shared a similar background and learning context at the macro level, guided by four research questions: (1) can the memory-based strategic vocabulary learning framework be employed to classify learners’ vocabulary learning strategies (VLS) satisfactorily?, (2) what strategies do Chinese university students use for learning vocabulary items?, (3) what are the learner clusters among Chinese university students regarding their use of VLS?, and (4) how are learners’ vocabulary learning approaches related to their language achievement? It is found that learners’ VLS use is very contextualised and subject to change due to many factors. A micro language-rich environment, where there is out of class reading and meaningful social interaction, is a key to high vocabulary achievement in an EFL context. The cluster analysis revealed a non-linear, complicated relationship between VLS use and vocabulary learning success. In addition, gender has a prominent impact on VLS use; however, the impact of learners’ discipline on VLS use is unclear and needs further investigation.

Globally, chronic diseases have a significant impact on health. The diagnosis of chronic diseases has seen extensive usage of machine learning techniques. Early disease detection and treatment lower the risk of increasing disease severity and, consequently, related mortality. The major goal of this research is to provide a technique that increases classification accuracy while also shortening computing time. This comparative research shows the impact of distinct model architectures and features on disease prediction accuracy in addition to assessing the advantages and disadvantages of each technique. These discoveries have implications for personalized healthcare, allowing medical professionals to select the best models for various chronic conditions. Additionally, this research can direct the creation of better forecasting technologies, as well as influence healthcare legislation and budget allocation. In our study comparative analysis of the state-of-the-art approaches has been presented. Using a hybrid model combination of CNN and RNN could be more beneficial. In conclusion, our comparison research improves our comprehension of the potential of deep machine learning for chronic disease prediction, highlighting the significance of adjusting model selection to certain disease types. To progress the field of chronic disease prediction, future research should concentrate on improving these models, and further explore their applicability across various and larger datasets.

This study presents model-free reinforcement learn ing methods for economic and ecological adaptive cruise control (Eco-ACC) of connected and autonomous electric vehicles. For model-free optimal control of Eco-ACC, we applied two reinforcement learning methods, Deep Q-Network (DQN) and Deep Deterministic Policy Gradient (DDPG), in which deep neural networks of actors and critics were trained using IPG CarMaker simulations. For performance demonstrations, the HWFET, US06, and WLTP Class 3b driving cycles were used to simulate the front vehicle, and the energy consumptions of the host vehicle and front vehicle were compared. In high-fidelity IPG CarMaker simulations, the proposed reinforcement learning- based Eco-ACC methods demonstrated approximately 3–5% and 10–14% efficiency improvements in highway and city-highway driving scenarios, respectively, when compared with the front vehicle. A video of the CarMaker simulation is available at https://youtu.be/DIXzJxMVig8.

This study explores the spatiotemporal expansion patterns of Airbnb hosts by applying the K-shape time-series clustering algorithm to host-level panel data. The analysis covers two periods—pre-pandemic (June 2014–December 2019) and post-pandemic (March 2021–July 2024)—while excluding the COVID-19 disruption phase. Hosts’ property expansion trajectories were examined over 24- and 36-month windows to identify recurring temporal patterns. The results reveal multiple forms of expansion and contraction behaviors, ranging from gradual and sustained growth to temporary decline and recovery. Comparing the two periods shows that post-pandemic host operations became more stable and less volatile. The study contributes to the literature on business expansion and professionalization in short-term rentals and provides practical insights for policymakers and platform managers aiming to foster sustainable and balanced market development.

Predicting traffic accidents is a challenging task because taking into account uncertainty in modeling traffic accidents is not trivial. To address these issues, this article develops a hybrid modeling pipeline combining unsupervised and supervised learning to predict the level of hazardous road sites and explore the causality of accidents by controlling unobserved heterogeneity issues effectively. Traffic accident data for Won-ju province, Korea, from 2020 to 2021, and external factors affecting traffic accidents, such as average travel speed and weather information, are combined based on road links. Through the modeling pipeline, a clustering technique is adopted to capture unobserved heterogeneous information among roads. Since traffic accident data contains a wide variety of categorical and hierarchical features, ensemble methods such as boosting techniques were applied to handle heterogeneity issues among these features. To explore the relationship between the accident and determinant factors, are adopted to interpret the results of machine learning models. Model-agnostic methods, however, generally provide results based on images, this study also added a process that extracts texts from images to overcome compatible issues with existing road safety management systems.

This study examined the impacts of utilizing deductive and inductive instructional approaches for teaching grammar in the instructional videos during the pre-class portion of a flipped learning class. More specifically, this study investigated the measured impacts of these instructional approaches on both effectiveness in the form of learning gains as well as student perceptions of both instructional options. The participants of this study consisted of 114 Korean university English-as-a-foreign-language students who were divided into two groups which had to watch a video utilizing either an inductive or deductive instructional approach before experiencing identical instructional approaches for the in-class portion of the flipped learning class. The analysis utilized Mann-Whitney U tests to compare both the learning gains and the learners’ perceptions in both groups of students. In terms of results, deductive and inductive instructional approaches were found to be similarly effective in the short-term and long-term and across proficiency groups. However, the deductive instructional approach was found to be slightly more interesting and easier than the inductive instructional approach.

The availability of detailed data on customers’ online behaviors and advances in big data analysis techniques enable us to predict consumer behaviors. In the past, researchers have built purchase prediction models by analyzing clickstream data; however, these clickstream-based prediction models have had several limitations. In this study, we propose a new method for purchase prediction that combines information theory with machine learning techniques. Clickstreams from 5,000 panel members and data on their purchases of electronics, fashion, and cosmetics products were analyzed. Clickstreams were summarized using the ‘entropy’ concept from information theory, while ‘random forests’ method was applied to build prediction models. The results show that prediction accuracy of this new method ranges from 0.56 to 0.83, which is a significant improvement over values for clickstream-based prediction models presented in the past. The results indicate further that consumers’ information search behaviors differ significantly across product categories.