Research Output
filter by:
Articles
Publication Date: 2026
Scientific Reports (20452322)16(1)
This study uses advanced time-series forecasting and causal modelling techniques to examine long-term patterns in Australian road traffic fatalities. Four statistical approaches were assessed: Holt-Winters, Theta, TBATS, and Vector Autoregression, with each offering strengths across different forecasting horizons. TBATS provided the most reliable short-term predictions, while Vector Autoregression performed best for medium- and long-term projections. A causal analysis using a random-effects panel model identified several key contributors to fatal crash risk, including older age groups, remote and outer-regional settings, nighttime periods, and high-speed environments. In contrast, younger adults and single-vehicle crashes were associated with lower fatality likelihood. Overall, the results demonstrate the value of flexible time-series techniques and panel data methods for guiding evidence-based road safety policy, targeted interventions, and infrastructure planning. © The Author(s) 2025.
Publication Date: 2025
Case Studies on Transport Policy (22136258)20
In Australia, road crash injuries continue to be a serious public health issue. Machine learning is used in this study to analyse injury data from road crashes between 2011 and 2021 that was taken from the national hospitalized injury database. We investigate how the number of injuries and duration of stay for road users are affected by variables such as gender, age, seasonal variation, collision type, and location (urban vs. regional). Road safety measures are informed by patterns and relationships found in the data by machine learning models. Hospitalizations have been trending upward between 2011 and 2019, with a pause in 2020 due to COVID-19 lockdowns. In all categories, men sustain more injuries than women, though the number varies according to age and geography. The type of road user also affects collision patterns. The time-series projections demonstrate that the goal of zero fatalities in 2050 will not be achieved under the business-as-usual scenario. The findings highlight the necessity of focused interventions predicated on collision trends and demographics. This includes better infrastructure design, increased surveillance, and customized safety measures. © 2025 The Author(s)
Publication Date: 2025
Machine Learning with Applications (26668270)22
Road traffic injuries continue to pose a significant public health challenge in Australia, with pedestrians representing one of the most vulnerable road user groups. Accurate prediction of injury severity, particularly fatal outcomes, is essential for improving road safety interventions and resource allocation. This study applies advanced machine learning techniques to predict pedestrian crash severity using national hospitalization and mortality data collected from 2011 to 2021. The analysis focuses on addressing class imbalance, a common issue in injury data by evaluating the impact of several data balancing methods, including SMOTE, ADASYN, Random Oversampling (ROS), and Threshold Moving. We implement and compare four supervised learning algorithms: Logistic Regression, Support Vector Machine (SVM), Decision Tree, and XGBoost. Model performance is assessed using F1-score and macro-accuracy, with a focus on the minority (fatality) class. Results show that XGBoost combined with Threshold Moving achieves the highest performance, yielding an F1-score of 72% for fatality classification and a macro-accuracy of 84%. Additionally, feature importance analysis using SHAP values reveals age, gender, road user type, and crash location as key predictors of injury severity. The study highlights the critical role of data balancing strategies in enhancing predictive accuracy for rare but high-impact outcomes. These findings provide actionable insights for transport authorities and policymakers seeking to develop data-driven, targeted safety measures to protect pedestrians and reduce the severity of crash outcomes. © 2025 The Author(s).
Publication Date: 2025
Injury (00201383)56(3)
Despite substantial progress in road safety, road traffic fatalities (RTFs) continue to be a persistent issue in Australia. This study aims to forecast RTFs trends up to 2050 by analyzing factors such as geographic location, age, gender, speed limits, and time of occurrence. Utilizing historical data from 1989 to 2024, fatalities were categorized by road user type, demographics, and day of the week. The Facebook Prophet time series model, incorporating categorical variables like region, age, and speed limits, was employed to predict future trends. The analysis reveals significant regional disparities in fatality reduction rates, with some areas lagging others. Gender-specific forecasts indicate a sharper decline in male fatalities compared to females, while projections highlight persistent risks for older drivers. Additionally, highways with higher speed limits are expected to see a substantial decrease in fatalities. These insights emphasize the need for targeted interventions in areas with slower reductions and high-risk demographic groups, aiding policymakers in refining safety measures, enforcing speed limits, and enhancing public awareness campaigns. © 2025 The Author(s)
Publication Date: 2024
pp. 291-296
The increase of traffic in Internet-of multimedia Things networks leads to additional load on servers; therefore, this paper focuses on server load balancing in multimedia Internet-of-Things networks. Software-defined networking technology has been used to achieve load balancing in these networks, as software-defined networks with new features have improved load balancing in multimedia Internet-of-Things networks. In this study, the short-term and long-term recurrent neural network algorithm is used to predict the server load, and then a fuzzy system is used to accurately determine the server levels. Also, this article saves energy and also reduces server overhead. © 2024 IEEE.
Publication Date: 2024
pp. 172-176
The Internet of Multimedia Things (IoMT) is an evolution of the IoT aimed at delivering multimedia streams as part of its realization. The IoMT is becoming increasingly appealing. Traditional computer network architectures are not robust enough to accommodate the rapid growth of IoMT networks. As a result, software-defined networks (SDNs) are utilized, which can centrally provide an overview of all network resources. SDNs enable advanced management by separating the control layer from the data layer. They introduce new capabilities to enhance load balancing. This research focuses on simultaneous load balancing between two key elements: servers and controllers. This approach allows us to prevent simultaneous issues in different network sections. Additionally, this approach employs long short-term memory prediction for forecasting the load on servers and controllers and a fuzzy system for distributing the load among servers and domains. Simulation results indicate that the proposed approach is highly effective in appropriately distributing load among servers in each network domain. The findings enable us to manage and optimize software-defined IoT networks more accurately using the proposed approach. This improves the quality of service provided to users and contributes to cost reduction and increased productivity. © 2024 IEEE.
Publication Date: 2023
pp. 158-162
Personalized QoE has significant implications for businesses in terms of customer satisfaction, loyalty, and revenue generation. By delivering experiences tailored to individual users, businesses can build stronger relationships, improve customer retention, and gain a competitive edge in the marketplace. In this paper, we have attempted to use a clustering-based approach to enhance personalized QoE assessment via personalized federated learning technique. To achieve this, first, we classify users to different clusters, based on some user-related QoE influencing factors. Second, we employ independent personalized federated learning QoE predictors in clusters to assess the QoE level of the service. We conducted some experiments to compare the performance of our method to the traditional personalized federated learning based QoE assessment approach. The results demonstrate that the proposed approach increases the accuracy of QoE evaluations by about 16% in average. © 2023 IEEE.
Publication Date: 2020
International Journal of Ad Hoc and Ubiquitous Computing (17438225)34(1)pp. 35-44
Wireless sensor networks have shown to be a promising technology for industrial automation in which continuous monitoring is a critical requirement. Deploying an energy-aware sensor network permits increasing the network lifetime and prolonging the monitoring operation. IEEE 802.15.4e and RPL have been used as de-facto protocols at the access and network layer in low power and low-rate wireless networks. Specifically, the time slotted channel hopping (TSCH) of 802.15.4e has been designed to provide a reliable access method in low power and lossy networks. More importantly, the combination of TSCH and RPL facilitates providing load-balancing together with energy-saving in such networks. This paper proposes schedule aware RPL (SA-RPL) which aims at prolonging the network lifetime while improving load balancing. It periodically collects scheduling matrix information form TSCH to compute a new measure for selecting the next hop at the network layer. More precisely, the parent with minimum number of occupied cells is more likely to be chosen as the preferred parent. To evaluate the performance of SA-RPL, we modified a distributed management scheme already developed in NS2 simulator. Simulation results show that SA-RPL, compared with other methods, prolongs the network lifetime up to two times and achieves a more uniform energy consumption distribution without decreasing other performance metrics. © 2020 Inderscience Enterprises Ltd.
Publication Date: 2016
Multimedia Tools and Applications (13807501)75(2)pp. 903-918
Measuring end user Quality of Experience (QoE) is currently performed by subjective or objective standard methods each with its own deficiencies. The subjective quality assessment is laboratory based, costly and offline; while, the objective estimation of user satisfaction is obtained through a static manner, not directly related to end user contentment. The attempt is made here to measure user QoE based on an online, user-aware and non-intrusive method. This is investigated by identifying the measurable objective indicators of user satisfaction/dissatisfaction and assigning them to the subjective nature of QoE concept. Proposing an architectural model, the extent of modalities for implicit sensing of the user QoE is explored with respect to the real-time measurement of her/his experiences. The vocal and interactional signs of VoIP service users on their smartphone devices are applied to estimate their satisfaction/dissatisfaction levels as a case study. The obtained results are compared to the users’ self-report in order to evaluate the accuracy of this proposed method. © 2014, Springer Science+Business Media New York.
Publication Date: 2014
Wireless Personal Communications (1572834X)79(3)pp. 2155-2170
Recently user quality of experience (QoE) is employed in evaluating end user satisfaction in communications systems. Generally, current approaches for QoE assessment are obtrusive, laboratory based and offline. Estimation of user satisfaction in static manner based on mean opinion score is not directly related to instantaneous individual end user contentment. In this paper, based on correlations between user’s physiological signals and her/his feelings about the service quality, a non-intrusive and user centric QoE assessment system for voice communications is developed. The findings of this study indicate that the emotional patterns in response to the changes in channel quality can be adapted to estimate the level of satisfaction in a QoE assessment system in a live manner. Based on experimental results, two categories of users are identified: sensitive and insensitive towards quality degradations. The results indicate that for the sensitive users, our non-intrusive subjective quality assessment method outperforms ITU-T P.563 standard with respect to root mean square error; while, the results are much better among the insensitive users. © 2014, Springer Science+Business Media New York.
