Browsing by Author "Zhao X"

Now showing 1 - 12 of 12
  • Thumbnail Image
    Item
    A comparative analysis of global optimization algorithms for surface electromyographic signal onset detection
    (Elsevier Inc, 2023-09) Alam S; Zhao X; Niazi IK; Ayub MS; Khan MA
    Surface Electromyography (sEMG) is a technique for measuring muscle activity by recording electrical signals from the surface of the body. It is widely used in fields such as medical diagnosis, human–computer interaction, and sports injury rehabilitation. The detection of the onset and offset of muscle activation is a longstanding challenge in sEMG analysis. This study pioneers the implementation, configuration, and evaluation of Particle Swarm Optimization (PSO) against other optimization algorithms for sEMG signal detection, including Genetic algorithms (GA), Simulated Annealing (SA), Ant Colony Optimization (ACO), and Tabu Search (TS). The results show that the PSO algorithm achieves the highest median accuracy and F1-Score and is the fastest among the selected algorithms but has lower stability compared to Genetic algorithms and Ant colony optimization. The design and value of the cost function had a significant impact on the results, with optimal results obtained when the cost value was between 0.1203 and 0.1384. The use of these algorithms improved detection efficiency and reduced the need for manual parameter adjustment. To the best of our knowledge, no published studies have utilized Simulated Annealing, Ant colony optimization, and Tabu search meta-heuristic algorithms to detect sEMG signal onsets.
  • Thumbnail Image
    Item
    A highway vehicle routing dataset during the 2019 Kincade Fire evacuation.
    (Springer Nature Limited, 2022-10-07) Xu Y; Zhao X; Lovreglio R; Kuligowski E; Nilsson D; Cova TJ; Yan X
    As the threat of wildfire increases, it is imperative to enhance the understanding of household evacuation behavior and movements. Mobile GPS data provide a unique opportunity for studying evacuation routing behavior with high ecological validity, but there are little publicly available data. We generated a highway vehicle routing dataset derived from GPS trajectories generated by mobile devices (e.g., smartphones) in Sonoma County, California during the 2019 Kincade Fire that started on October 23, 2019. This dataset contains 21,160 highway vehicle routing records within Sonoma County from October 16, 2019 to November 13, 2019. The quality of the dataset is validated by checking trajectories and average travel speeds. The potential use of this dataset lies in analyzing and modeling evacuee route choice behavior, estimating traffic conditions during the evacuation, and validating wildfire evacuation simulation models.
  • Thumbnail Image
    Item
    A Semi-automatic Diagnosis of Hip Dysplasia on X-Ray Films
    (Frontiers Media S.A., 2020-12-17) Yang G; Jiang Y; Liu T; Zhao X; Chang X; Qiu Z; Gao X
    Background: Diagnosis of hip joint plays an important role in early screening of hip diseases such as coxarthritis, heterotopic ossification, osteonecrosis of the femoral head, etc. Early detection of hip dysplasia on X-ray films may probably conduce to early treatment of patients, which can help to cure patients or relieve their pain as much as possible. There has been no method or tool for automatic diagnosis of hip dysplasia till now. Results: A semi-automatic method for diagnosis of hip dysplasia is proposed. Considering the complexity of medical imaging, the contour of acetabulum, femoral head, and the upper side of thigh-bone are manually marked. Feature points are extracted according to marked contours. Traditional knowledge-driven diagnostic criteria is abandoned. Instead, a data-driven diagnostic model for hip dysplasia is presented. Angles including CE, sharp, and Tonnis angle which are commonly measured in clinical diagnosis, are automatically obtained. Samples, each of which consists of these three angle values, are used for clustering according to their densities in a descending order. A three-dimensional normal distribution derived from the cluster is built and regarded as the parametric model for diagnosis of hip dysplasia. Experiments on 143 X-ray films including 286 samples (i.e., 143 left and 143 right hip joints) demonstrate the effectiveness of our method. According to the method, a computer-aided diagnosis tool is developed for the convenience of clinicians, which can be downloaded at http://www.bio-nefu.com/HIPindex/. The data used to support the findings of this study are available from the corresponding authors upon request. Conclusions: This data-driven method provides a more objective measurement of the angles. Besides, it provides a new criterion for diagnosis of hip dysplasia other than doctors' experience deriving from knowledge-driven clinical manual, which actually corresponds to very different way for clinical diagnosis of hip dysplasia.
  • Thumbnail Image
    Item
    Analyzing Risk Perception, Evacuation Decision and Delay Time: A Case Study of the 2021 Marshall Fire in Colorado
    (Elsevier B.V., 2023-12-11) Forrister A; Kuligowski ED; Sun Y; Yan X; Lovreglio R; Cova TJ; Zhao X
    Climate change is increasing the threat of wildfires to populated areas, especially those within the wildland-urban interface (WUI). The 2021 Marshall fire forced the evacuation of over 30,000 people in Boulder, Jefferson and Adams Counties in Colorado, US. To improve our understanding of wildfire evacuation response, we surveyed individuals affected by the Marshall fire to analyze their evacuation decisions and resulting behavior. We used linear and logistic regression models to determine the factors influencing individuals’ risk perceptions, their decisions to evacuate or stay, and the associated evacuation delay times. We found higher levels of risk perception at the time of the evacuation decision were associated with higher levels of pre-fire perceived risk, having mid-level household income, the receipt of fire cues and having a medical condition. Increased pre-event risk perception increased the likelihood of evacuating, along with gender (female-identified), being aged between 55 and 64 years, and having a higher household income. On the other hand, having a prior awareness of wildfires had a negative effect on evacuation likelihood. Additionally, having previous experience with fire damage, owning their home, having a larger household size and being alerted later in the fire event reduced the delay time; whereas engaging in preparation activities and having children in the home led to longer delay times. These research findings can be used by emergency managers to better prepare WUI communities for future wildfire events.
  • Thumbnail Image
    Item
    Clustering by Search in Descending Order and Automatic Find of Density Peaks
    (IEEE, 2019-01-01) Liu T; Li H; Zhao X; Liang Q
    Clustering by fast search and find of density peaks published on journal Science in 2014 is a density-based clustering technique, which is not only unnecessary to determine the number of clusters in advance, but also able to recognize the clusters of arbitrary shapes. Due to a manual selection of clustering centers on a decision graph, samples which belong to one cluster may be assigned to two or more clusters and vice versa. On assumption that boundary points which keep comparable densities with cluster centers should be regarded as inner points, we make a new method which not only can find all possible clusters automatically but also can combine those with similarities simultaneously to obtain the final clusters. Unlike clustering by fast search and find of density peaks, we only focus on densities with discarding the relative metric which measures the minimum distance between a cluster center and a point with a higher density. Qualitative and quantitative experimental results on sufficient datasets demonstrate the effectiveness of our method.
  • Thumbnail Image
    Item
    Evolutionary Game and Simulation of Green Housing Market Subject Behavior in China.
    (John Wiley and Sons, 2022-04-05) Qian Y; Yu M; Wang T; Yuan R; Feng Z; Zhao X; Fu H
    In China, driven by the national "3060" double carbon targets (i.e., reaching peak carbon emissions by 2030 and carbon neutrality by 2060), green housing has become one of the major fields to reduce carbon emissions, facilitating the achievement of the double carbon targets. Promoting the growth of green housing is an important way for the real estate industry to achieve low-carbon transformation and improve the quality of housing. Meanwhile, the construction industry also can benefit from green housing to achieve its energy conservation and emission reduction targets. Therefore, it is critical to boost and maintain the sustainable growth of the green housing market in China. However, the literature has not focused attention on the market behavior of the green housing market in China. This study proposes a tripartite evolutionary game model to investigate the subject behavior of the green housing market in China. This model consists of three major subjects in a green housing market: developers, consumers, and governments. Based on this model, this study analyzes the stability of the strategy options for each stakeholder and identifies the stable conditions of strategy portfolios to reach the equilibrium points of the game system. The validity of the proposed tripartite evolutionary game model is tested through the simulation of the impacts from various factors on system evolution. According to the impacts of factors and the stable conditions of strategies, this paper puts forward relevant policy suggestions for the healthy and sustainable growth of China's green housing market.
  • Thumbnail Image
    Item
    Influences of nitrogen input forms and levels on phosphorus availability in karst grassland soils
    (Frontiers Media S.A., 2024-01-18) Zhou J; Yang F; Zhao X; Gu X; Chen C; Chen J
    The availability of soil phosphorus (P), a crucial nutrient influencing plant productivity and ecosystem function, is impacted by continuously increasing nitrogen (N) enrichment, which changes the soil P cycle. The effect of varying forms of N input on soil P dynamics in P-limited karst grassland ecosystems remains unclear. To address this knowledge gap, we conducted a greenhouse experiment to explore the effects of various forms of N addition [Ca(NO3)2, NH4Cl, NH4NO3, Urea] on soil P fractions in these ecosystems, applying two levels (N1: 50 mg N kg−1soil, N2: 100 mg N kg−1soil) of N input in two soils (yellow soil, limestone soil). Results indicated that P fractions in both soil types were significantly affected by N additions, with yellow soil demonstrating a higher sensitivity to these additions, and this effect was strongly modulated by the form and level of N added. High N addition, rather than low N, significantly affect the P fractions in both soil types. Specially, except for Ca(NO3)2, high N addition significantly increased the available P in both soils, following the order: Urea and NH4NO3 > NH4Cl > Ca(NO3)2, and decreased NaHCO3-Pi in both soils. High N addition also significantly reduced NaOH-Po and C.HCl-Po fractions in yellow soil. Additionally, the response of root biomass and alkaline phosphatase activity in both soils to N input paralleled the trends observed in the available P fractions. Notably, changes in soil available P were strongly correlated with plant root biomass and soil alkaline phosphatase activity. Our study highlights that the N addition form significantly influences soil P availability, which is closely tied to plant root biomass and alkaline phosphatase activity. This finding underscores the importance of considering N input form to boost soil fertility and promote sustainable agriculture.
  • Thumbnail Image
    Item
    Modelling and interpreting pre-evacuation decision-making using machine learning
    (Elsevier BV, 2020-05) Zhao X; Lovreglio R; Nilsson D
    The behaviour of building occupants in the first stage of an evacuation can dramatically impact the time required to evacuate buildings. This behaviour has been widely investigated by scholars with a macroscopic approach fitting random distributions to represent the pre-evacuation time, i.e. time from noticing the first cue until deliberate movement. However, microscopic investigations on how building occupants respond to several social and environmental factors are still rare in the literature. This paper aims to leverage machine learning as a possible solution to investigate factors affecting building occupants' decision-making during pre-evacuation stage. In particular, we focus on applying interpretable machine learning to reveal the interactions among the input variables and to capture nonlinear relationships between the input variables and the outcome. As such, we use a well-established machine-learning algorithm—random forest—to model and predict people's emergency behaviour pre-evacuation. We then apply tools to interpret the black-box random forest model to extract useful knowledge and gain insights for emergency planning. Specifically, this algorithm is applied here to investigate the behaviour of 569 building occupants split between five unannounced evacuation drills in a cinema theatre. The results indicate that both social and environmental factors affect the probability of responding. Several independent variables, such as the time elapsed after the alarm has started and the decision-maker's group size, are presenting strong nonlinear relationships with the probability of switching to the response stage. Furthermore, we find interactions exist between the row number where the decision-maker sits and the number of responding occupants visible to her; the complex relationship between the outcome and these two variables can be visualized by using a two-dimensional partial dependence plot. An interesting finding is that a decision-maker is more sensitive to the proportion of responding occupants than the number of them; hence, the people sitting in the back are often responding more slowly than the people in the front.
  • Thumbnail Image
    Item
    Situational-aware multi-graph convolutional recurrent network (SA-MGCRN) for travel demand forecasting during wildfires
    (Elsevier B.V., 2024-09-10) Zhang X; Zhao X; Xu Y; Nilsson D; Lovreglio R
    Natural hazards, such as wildfires, pose a significant threat to communities worldwide. Real-time forecasting of travel demand during wildfire evacuations is crucial for emergency managers and transportation planners to make timely and better-informed decisions. However, few studies focus on accurate travel demand forecasting in large-scale emergency evacuations. To tackle this research gap, the study develops a new methodological framework for modeling highly granular spatiotemporal trip generation in wildfire evacuations by using (a) large-scale GPS data generated by mobile devices and (b) state-of-the-art AI technologies. Based on the travel demand inferred from the GPS data, we develop a new deep learning model, i.e., Situational-Aware Multi-Graph Convolutional Recurrent Network (SA-MGCRN), along with a model updating scheme to achieve real-time forecasting of travel demand during wildfire evacuations. The proposed methodological framework is tested using a real-world case study: the 2019 Kincade Fire in Sonoma County, CA. The results show that SA-MGCRN significantly outperforms all the selected state-of-the-art benchmarks in terms of prediction performance. Our finding suggests that the most important model components of SA-MGCRN are weekend indicator, population change, evacuation order/warning information, and proximity to fire, which are consistent with behavioral theories and empirical findings. SA-MGCRN can be directly used in future wildfire events to assist real-time decision-making and emergency management.
  • Thumbnail Image
    Item
    Social vulnerabilities and wildfire evacuations: A case study of the 2019 Kincade fire
    (Elsevier B.V., 2024-05-31) Sun Y; Forrister A; Kuligowski ED; Lovreglio R; Cova TJ; Zhao X
    Vulnerable populations (e.g., populations with lower income or disabilities) are disproportionately impacted by natural hazards like wildfires. It is crucial to develop equitable and effective evacuation strategies to meet their unique needs. While existing studies offer valuable insights, we need to improve our understanding of how vulnerabilities affect wildfire evacuation decision-making, as well as how this varies spatially. The goal of this study is to conduct an in-depth analysis of the impacts of social vulnerabilities on aggregated evacuation decisions, including evacuation rates, delay in departure time, and evacuation destination distance by leveraging large-scale GPS data generated by mobile devices. Specifically, we inferred evacuation decisions at the level of the census block group, a geographic unit defined by the U.S. Census, utilizing GPS data. We then employed ordinary least squares and geographically weighted regression models to investigate the impacts of social vulnerabilities on evacuation decisions. We also used Moran's I to test if these impacts were consistent across different block groups. The 2019 Kincade Fire in Sonoma County, California, was used as the case study. The impacts of social vulnerabilities on evacuation rates show significant spatial variations across block groups, whereas their effects on the other two decision types do not. Additionally, unemployment, a factor under-explored in previous studies, was identified as contributing to both an increased delay in departure time and a reduction in destination distance of evacuees at the aggregate level. Furthermore, upon comparing the significant factors across different models, we observed that some of the vulnerabilities contributing to evacuation rates for all residents differed from those affecting the delay in departure time and destination distance, which only applied to evacuees. These new insights can guide emergency managers and transportation planners to enhance equitable wildfire evacuation planning and operations.
  • Thumbnail Image
    Item
    The relationship between hair metabolites, air pollution exposure and gestational diabetes mellitus: A longitudinal study from pre-conception to third trimester.
    (Frontiers Media S.A., 2022-12-02) Chen X; Zhao X; Jones MB; Harper A; de Seymour JV; Yang Y; Xia Y; Zhang T; Qi H; Gulliver J; Cannon RD; Saffery R; Zhang H; Han T-L; Baker PN; Zhou N
    BACKGROUND: Gestational diabetes mellitus (GDM) is a metabolic condition defined as glucose intolerance with first presentation during pregnancy. Many studies suggest that environmental exposures, including air pollution, contribute to the pathogenesis of GDM. Although hair metabolite profiles have been shown to reflect pollution exposure, few studies have examined the link between environmental exposures, the maternal hair metabolome and GDM. The aim of this study was to investigate the longitudinal relationship (from pre-conception through to the third trimester) between air pollution exposure, the hair metabolome and GDM in a Chinese cohort. METHODS: A total of 1020 women enrolled in the Complex Lipids in Mothers and Babies (CLIMB) birth cohort were included in our study. Metabolites from maternal hair segments collected pre-conception, and in the first, second, and third trimesters were analysed using gas chromatography-mass spectrometry (GC-MS). Maternal exposure to air pollution was estimated by two methods, namely proximal and land use regression (LUR) models, using air quality data from the air quality monitoring station nearest to the participant's home. Logistic regression and mixed models were applied to investigate associations between the air pollution exposure data and the GDM associated metabolites. RESULTS: Of the 276 hair metabolites identified, the concentrations of fourteen were significantly different between GDM cases and non-GDM controls, including some amino acids and their derivatives, fatty acids, organic acids, and exogenous compounds. Three of the metabolites found in significantly lower concentrations in the hair of women with GDM (2-hydroxybutyric acid, citramalic acid, and myristic acid) were also negatively associated with daily average concentrations of PM2.5, PM10, SO2, NO2, CO and the exposure estimates of PM2.5 and NO2, and positively associated with O3. CONCLUSIONS: This study demonstrated that the maternal hair metabolome reflects the longitudinal metabolic changes that occur in response to environmental exposures and the development of GDM.
  • Thumbnail Image
    Item
    Variable Selection from Image Texture Feature for Automatic Classification of Concrete Surface Voids.
    (Hindawi Limited, 2021-03-08) Zhao Z; Liu T; Zhao X; Haber RE
    Machine learning plays an important role in computational intelligence and has been widely used in many engineering fields. Surface voids or bugholes frequently appearing on concrete surface after the casting process make the corresponding manual inspection time consuming, costly, labor intensive, and inconsistent. In order to make a better inspection of the concrete surface, automatic classification of concrete bugholes is needed. In this paper, a variable selection strategy is proposed for pursuing feature interpretability, together with an automatic ensemble classification designed for getting a better accuracy of the bughole classification. A texture feature deriving from the Gabor filter and gray-level run lengths is extracted in concrete surface images. Interpretable variables, which are also the components of the feature, are selected according to a presented cumulative voting strategy. An ensemble classifier with its base classifier automatically assigned is provided to detect whether a surface void exists in an image or not. Experimental results on 1000 image samples indicate the effectiveness of our method with a comparable prediction accuracy and model explicable.