The application of machine learning in multi sensor data fusion for activity recognition in mobile device space

In recent years, mobile devices have gained increasing development with stronger computation capability and larger storage space. Some of the computation-intensive machine learning tasks can now be run on mobile devices. To exploit the resources available on mobile devices and preserve personal privacy, the concept of client-based machine learning has been proposed. It leverages the users’ local hardware and local data to solve machine learning sub-problems on mobile devices and only uploads computation results rather than the original data for the optimization of the global model. Such an architecture can not only relieve computation and storage burdens on servers but also protect the users’ sensitive information. Another benefit is the bandwidth reduction because various kinds of local data can be involved in the training process without being uploaded. In this article, we provide a literature review on the progressive development of machine learning from server based to client based. We revisit a number of widely used server-based and client-based machine learning methods and applications. We also extensively discuss the challenges and future directions in this area. We believe that this survey will give a clear overview of client-based machine learning and provide guidelines on applying client-based machine learning to practice.

Embedded systems technology is undergoing a phase of transformation owing to the novel advancements in computer architecture and the breakthroughs in machine learning applications. The areas of applications of embedded machine learning (EML) include accurate computer vision schemes, reliable speech recognition, innovative healthcare, robotics, and more. However, there exists a critical drawback in the efficient implementation of ML algorithms targeting embedded applications. Machine learning algorithms are generally computationally and memory intensive, making them unsuitable for resource-constrained environments such as embedded and mobile devices. In order to efficiently implement these compute and memory-intensive algorithms within the embedded and mobile computing space, innovative optimization techniques are required at the algorithm and hardware levels. To this end, this survey aims at exploring current research trends within this circumference. First, we present a brief overview of compute intensive machine learning algorithms such as hidden Markov models (HMM), k-nearest neighbors (k-NNs), support vector machines (SVMs), Gaussian mixture models (GMMs), and deep neural networks (DNNs). Furthermore, we consider different optimization techniques currently adopted to squeeze these computational and memory-intensive algorithms within resource-limited embedded and mobile environments. Additionally, we discuss the implementation of these algorithms in microcontroller units, mobile devices, and hardware accelerators. Conclusively, we give a comprehensive overview of key application areas of EML technology, point out key research directions and highlight key take-away lessons for future research exploration in the embedded machine learning domain.

Machine learning applications for anomaly detection in Smart Water Metering Networks: A systematic review

Chemical heuristics have been fundamental to the advancement of chemistry and materials science. These heuristics are typically established by scientists using knowledge and creativity to extract patterns from limited datasets. Machine learning offers opportunities to perfect this approach using computers and larger datasets. Here, we discuss the relationships between traditional heuristics and machine learning approaches. We show how traditional rules can be challenged by large-scale statistical assessment and how traditional concepts commonly used as features are feeding the machine learning techniques. We stress the waste involved in relearning chemical rules and the challenges in terms of data size requirements for purely data-driven approaches. Our view is that heuristic and machine learning approaches are at their best when they work together.

Introduction: The introduction of Clinical Information Systems (CIS) to manage and store patient information has deeply modified the workflow of caregivers. As the information is no longer stored on a physical medium, caregivers depend on the presence of a computer to access patient data. This constraint has built up an invisible barrier between the source of information and the place of its use. The popularisation of mobile devices offers an oportunity to break this barrier by providing a ubiquitous access to information. The use of a mobile platform merges the advantages of a modern CIS and the mobility offered by paper. We have developed an application on a mobile device that nurses can carry along on all their visits. It allows the information gathered during their visits to be entered directly into the system. Method: The tool we present aims to manage the daily interventions of nurses. The interventions concern all kinds of treatments provided by nurses to patients. In the current organisation, nurses start their day by printing out all the daily interventions and use it as a procedure to follow. Every time a nurse performs an intervention, she takes a note indicating that the task has been done properly. When they have some spare time, nurses enter all the gathered information in the CIS. Once nurses are equipped with mobile devices, paper becomes useless. Consequently, the process is simplified. Replacing the paper by a mobile device application is not straightforward. In order to access patient information, the tool must be linked to the existing CIS of the hospital without creating too strong dependencies. Moreover, the mobile interface must be adapted to the specificities of the mobile device such as a smaller display size, a reduced computational power, a tactile-based interaction and a real-time usage. Results: In order to link the mobile application to the existing CIS, we relied on a Service Oriented Architecture (SOA). It was mandatory to think of a software architecture that would not create a dependency with any legacy system. Thus, we defined a gateway server providing centralised access for the mobile application to any required information to or from the CIS. Integrating any mobile application would only require integrating this bridge. The gateway server also clearly separates the services that are available remotely from the ones proposed as usual Web services. The developed application not only allows the validation of the interventions and the input of measures, but it also offers a simple way to define the context in which nurses are working. This context is necessary to identify the specific list of interventions for the patients under the caregiver’s responsibility. The current CIS of the HUG includes a module that provides information concerning the interventions. Its interface is adjusted for a personal computer screen. Since caregivers are accustomed to this organisation of the information, we aimed to develop an interface on a mobile device that guarantees a certain level of homogeneity with the existing PC application. However, a mobile device’s screen is of limited size, therefore the interface has been completely rebuilt to be adapted to the new display. To deal with the limited display size, we rely on hierarchy and indexicality principles. For instance, based on the hypothesis that all tasks belonging to one category and planned at the same time can be regrouped under a single item; we regroup the interventions of similar top level in a common item. Moreover, to minimise user manipulations, relevant items are automatically moved to the top. Conclusion: The current daily workflow of nurses cannot be managed consistently using a unique tool. The need to switch from paper to computer induces work overload and is a source of errors. Therefore, we have developed a program running on a mobile device that nurses can carry during their visits. This program allows to manage nurses' daily interventions in real-time. The development of such a tool has required a complete reconstruction of the classical personal computer interface. This new interface takes into account a smaller display size, an innovative interaction paradigm, and real-time constraints.

The oil and gas industry faces constant pressure to enhance operational efficiency, reduce costs, and minimize equipment downtime. Machine learning (ML) applications in predictive maintenance have emerged as a transformative approach to achieving these goals. This review explores the role of machine learning in predictive maintenance, highlighting its potential to revolutionize maintenance strategies and improve asset management across the industry. Predictive maintenance leverages advanced algorithms to analyze historical and real-time data from equipment and sensors, enabling the identification of patterns and anomalies that precede equipment failures. By utilizing techniques such as supervised learning, unsupervised learning, and reinforcement learning, organizations can forecast equipment malfunctions and schedule maintenance activities proactively, thereby reducing unexpected downtimes and extending asset lifecycles. The implementation of ML in predictive maintenance provides several key benefits. Firstly, it enhances operational efficiency by optimizing maintenance schedules and minimizing unplanned outages, which are critical in the capital-intensive oil and gas sector. Secondly, it enables cost savings through more efficient resource allocation and reduced labor costs associated with reactive maintenance strategies. Thirdly, machine learning algorithms can continuously learn from new data, refining their predictive capabilities and improving accuracy over time. Several case studies illustrate the successful application of machine learning in predictive maintenance within the oil and gas industry. For example, ML models have been employed to predict pump failures, optimize drilling operations, and improve pipeline integrity monitoring. These applications not only lead to significant financial savings but also enhance safety by reducing the risk of catastrophic failures. In conclusion, machine learning applications in predictive maintenance represent a crucial advancement for the oil and gas industry. By harnessing the power of data-driven insights, organizations can enhance operational efficiency, reduce costs, and ultimately drive sustainable growth. This review emphasizes the transformative potential of machine learning in predictive maintenance, establishing it as a key strategy for success in the ever-evolving oil and gas landscape.

Edge computing enables smart IoT-based systems via concurrent and continuous execution of latency-sensitive machine learning (ML) applications. These edge-based machine learning systems are often battery-powered (i.e., energy-limited). They use heterogeneous resources with diverse computing performance (e.g., CPU, GPU, and/or FPGA) to fulfill the latency constraints of ML applications. The challenge is to allocate user requests for different ML applications on the Heterogeneous Edge Computing Systems (HEC) with respect to both the energy and latency constraints of these systems. To this end, we study and analyze resource allocation solutions that can increase the on-time task completion rate while considering the energy constraint. Importantly, we investigate edge-friendly (lightweight) multi-objective mapping heuristics that do not become biased toward a particular application type to achieve the objectives; instead, the heuristics consider "fairness" across the concurrent ML applications in their mapping decisions. Performance evaluations demonstrate that the proposed heuristic outperforms widely-used heuristics in heterogeneous systems in terms of the latency and energy objectives, particularly, at low to moderate request arrival rates. We observed 8.9% improvement in on-time task completion rate and 12.6% in energy-saving without imposing any significant overhead on the edge system.

Purpose Although there are studies analyzing barriers related to new technological concepts, it turns out that there are only a few studies on barriers to machine learning (ML) applications, and none of them consider the implications for smart logistics. Therefore, the purpose of this study is to reveal and analyze the barriers to ML applications in smart logistics from both industry and academic perspectives. Design/methodology/approach To achieve this aim, first, various barriers to smart logistics activities based on the Industry 4.0 perspective are identified. Later, the relative importance of these barriers critical to the success of smart logistics activities is determined. Finally, the interval-valued fuzzy (IVF) DEMATEL method is used to analyze the cause-and-effect relationship between each barrier based on industry and academic perspective. Findings Eleven barriers related to ML applications in smart logistics were evaluated by seven experts who are working in different positions. Results show that, the most crucial cause-and-effect barriers are integration and connection problems with value chain/network systems (B6), requirements of adapting new infrastructures (B11) and lack of transparency, safety and security (B3). Originality/value There is no study about determining barriers with merging smart logistics activities with the Industry 4.0 perspective. It is expected that the results of this study will contribute to the use of ML in the logistics sector by revealing significant concepts to which businesses should pay attention to prevent these barriers and by suggesting practical solutions to these problems.

This study investigates the scientific development of Artificial Intelligence (AI) and Machine Learning (ML) applications in Human Resource Management (HRM) through bibliometric analysis. To this end, 522 academic publications indexed in the Scopus database between 2020 and 2024 were analyzed using the R-based bibliometrix package and VOSviewer software. Descriptive analysis, scientific productivity metrics, and content analysis techniques were employed. The findings revealed three main patterns. First, research on AI and ML applications in HRM has grown significantly—particularly between 2022 and 2024—driven by post-pandemic digital transformation. Second, India, China, and the USA lead in research output, while the UK and France demonstrate strong citation impact, indicating a globally expanding research ecosystem. Third, the thematic focus of research is shifting from technical infrastructure toward more human-centered and ethical dimensions. Additionally, keyword co-occurrence network analysis identified three major thematic clusters: HRM functions, AI applications, and machine learning analytics, highlighting the field’s interdisciplinary nature. Compared to the previous studies, this research provides a more comprehensive bibliometric analysis of AI and ML applications in HRM. It is the first extensive study to map the intellectual evolution of the field from a multidisciplinary perspective. Furthermore, it charts research trends and collaboration networks, revealing a shift from technical implementations to strategic integration. In conclusion, this analysis offers new insights to the literature by illustrating the technological evolution in HRM and highlighting the growing significance of cutting-edge approaches such as AI and ML, reaffirming the field as a timely and impactful area of research.

This paper examines the enhancement of security measures for the Internet of Things (IoT) systems through the application of Machine Learning (ML) techniques. As the number of IoT devices continues to rise, ensuring their security has become increasingly critical, given that conventional methods frequently struggle to identify advanced threats. This study explores the implementation of several ML algorithms, including Random Forest (RF), Decision Trees (DT), Support Vector Machines (SVM), and Convolutional Neural Networks (CNN), to identify anomalies and intrusions within IoT networks. By conducting a comprehensive review of existing research and experiments, it highlights the effectiveness of ML in enhancing IoT security, with high detection rates for various threats, including botnet attacks, Denial-of-Service (DoS) and Distributed Denial-of-Service (DDoS) incidents, and intrusion attempts. DoS/DDoS attacks and many types of botnets are the most devastating attacks that have been spreading for a long time, and they are still branching out in new ways against IoT networks. They can damage IoT services and prevent these services from being used by legitimate users. Therefore, securing IoT networks becomes a significant concern. The proposed model is used to increasingly monitor network traffic for any deviations from standard patterns IoT networks. This paper also stresses the necessity of utilising suitable datasets and feature selection techniques to enhance the efficacy of ML models. To train our model, we have utilized a dataset called the IoT23 dataset, which is one of the most recent datasets that has many IoT scenarios and anomalous activities. Further-more, we utilised two types of feature selection algorithms, the Correlation-based Feature Selection (CFS) algorithm and the Genetic Algorithm (GA), and then we compared the results of these algorithms when training our model. The best performances were obtained with DT and RF classifiers when they were trained with features selected by CFS However, for training and testing times metrics, DT performance was superior across both feature selection methods.

The current review examines the state of knowledge and research on machine learning (ML) applications in horticultural production and the potential for predicting fresh produce losses and waste. Recently, ML has been increasingly applied in horticulture for efficient and accurate operations. Given the health benefits of fresh produce and the need for food and nutrition security, efficient horticultural production and postharvest management are important. This review aims to assess the application of ML in preharvest and postharvest horticulture and the potential of ML in reducing postharvest losses and waste by predicting their magnitude, which is crucial for management practices and policymaking in loss and waste reduction. The review starts by assessing the application of ML in preharvest horticulture. It then presents the application of ML in postharvest handling and processing, and lastly, the prospects for its application in postharvest loss and waste quantification. The findings revealed that several ML algorithms perform satisfactorily in classification and prediction tasks. Based on that, there is a need to further investigate the suitability of more models or a combination of models with a higher potential for classification and prediction. Overall, the review suggested possible future directions for research related to the application of ML in postharvest losses and waste quantification.

A cost effective, easily scalable, and application independent FPGA cluster co-processing platform for machine learning (ML) applications is proposed. The work in this paper focuses on delivering an economical platform for the researchers and developers without the knowledge of FPGAs who seek budget friendly solutions to increase the performance of their ML applications without relying on the highly expensive solutions available in the today's FPGA market. The approach consists of two main parts, with a CPU based host machine learning application and the FPGA cluster sharing the heavy workload through an Ethernet protocol. An experiment was conducted using a perceptron layer implementation on the hardware and the test results show the execution time improvements due to the utilization of the proposed cluster approach. The solution proposed is easily scalable with any FPGA platform with an Ethernet support.

Data fusion and machine learning for industrial prognosis: Trends and perspectives towards Industry 4.0

Offloading footwear is a critical intervention for preventing and treating diabetic foot ulcers (DFUs), as it reduces plantar pressure (PP) and promotes healing. However, decision-making around its prescription remains fragmented, with variability in feature selection, limited personalization, and diverse evaluation approaches. While guidelines, knowledge-based systems, and machine learning (ML) applications have explored elements of decision-making, no clinical decision support system (CDSS) currently exists to provide feature-level prescriptions. We conducted a narrative review of 45 studies, including 12 guidelines/protocols, 25 knowledge-based systems, and eight ML applications. Studies published from inception to August 2025 were included. Studies were thematically analyzed by type of knowledge, decision logic, evaluation methods, and associated technologies. Guidelines emphasize PP thresholds (≤200 kPa or ≥25–30% reduction) but lack detailed outputs. Knowledge-based systems apply rule-based and sensor-driven logic, integrating PP monitoring, adherence tracking, and usability testing. ML applications introduce predictive classification, optimization, and generative models, achieving high computational accuracy but with limited explainability and clinical validation. Evaluation practices remain fragmented: biomechanical testing dominates protocols, usability and adherence assessments are common in knowledge-based systems, while ML studies emphasize technical accuracy without linking to long-term clinical outcomes. From this synthesis, we propose a five-component framework for CDSS development that discusses: (1) minimum viable dataset, (2) hybrid decision architecture combining rules, optimization, and explainable ML, (3) structured feature-level outputs, (4) continuous validation and evaluation, and (5) integration into clinical and telehealth workflows. The proposed framework presents a pathway to transform fragmented approaches into scalable, patient-centered CDSSs. Prioritizing interoperable datasets, explainable models, and outcome-driven evaluation will be critical for clinical adoption and improved DFU care.

Recently, smart phones such as the Galaxy S and IPhone, and tablet PCs such as the Galaxy Tab have been introduced and become widespread in the market, thus the environment for wireless Internet services has been improving. As the market has improved, consumer preferences for mobile devices and the relationship between these devices and wireless Internet services regarding consumer preferences have become a very important issue. However, although there have been a considerable number of studies on the demand for mobile devices, it is not easy to find previous studies on the relationship between consumer preferences for specific attributes of wireless Internet services and the demand for mobile devices. The main purpose of this study is to analyze which mobile device is more likely to be preferred or become the main channel for wireless Internet services and how the demand for each mobile device is influenced by consumer preferences for specific attributes of wireless broadband Internet services used on it, and to induce managerial implications based on the main findings. To derive the results on consumer preferences, conjoint analysis is used. The survey was conducted using 1,000 Korean consumers. In the survey, respondents are allowed to choose more than one hypothetical alternative representing a mobile device combined with a certain type of wireless broadband Internet service if they desired. To estimate the consumer preferences quantitatively, a multivariate probit model is used. According to the estimation results, owing to their portability, smart phones are more likely to become a major channel for wireless Internet services despite their drawbacks in terms of display size and performance. Portability is the most important factor affecting the choice of a mobile device when using wireless Internet services. Therefore, manufacturers of mobile devices for wireless Internet services should improve the overall functionality and performances of mobile devices without negatively affecting portability. However, at the same time, consumer preferences for specific attributes of wireless Internet services are found to affect the demand of each mobile device much differently. Therefore, the prediction that smart phones will be the prevalent mobile device for use of wireless Internet services can be somewhat changed, depending on which and how wireless Internet services are provided to each mobile device and how they affect the demand for each mobile device. As mentioned above, there are significant differences between mobile devices with regard to the effects of mobility, data transfer speed, type of contents or Internet interface for wireless Internet services. It is found that the mobility of wireless Internet service is highly valued for smaller devices, but not much for notebook PCs. In contrast, data transfer speed and similarities in the type of contents or Internet interface to those provided by fixed broadband Internet services become more important for larger mobile devices. Thus, it is necessary for wireless Internet service providers to differentiate and optimize their contents and Internet interfaces properly for various mobile devices. Additionally, it is reasonable that manufacturers of smart phones focus on mobility and portability (or convenience) rather than adding sophisticated functions or pursuing higher-level specifications. Further, for smart phones, for which mobility and convenience are important, wireless Internet service providers should develop and provide easy and convenient programs, rather than advanced programs that require high device performance. Lastly, with respect to individual demographics, young and highly educated people are more likely to be consumers of newly introduced mobile devices for using wireless Internet services. Therefore, these young and highly educated consumers should be considered a major target for manufacturers of mobile devices for wireless Internet services.