机器学习支持下的老年人步行环境满意度与适老环境改善策略研究OA北大核心CSTPCD
Machine learning analysis of seniors'satisfaction and improvement strategy of the walking environment
当前,我国正面临着严峻的人口老龄化挑战.鉴于步行的多重价值,营造适老社区步行环境已成为当务之急.本研究基于成都市1 012名老年人对步行环境的满意度问卷调查数据,采用随机森林模型确定了各个步行环境指标的重要性,并进行了重要性—绩效分析,明确了步行环境指标的改善优先级.研究结果显示,社区街景、生活服务设施可达性、步行道宽度和座椅可及性是影响老年人步行满意度最为显著的四个指标,而老年人对于十字路口间距和步行道连通性这两个指标的关注较为有限.座椅可及性、社区街景、商业中心可达性和休闲设施可达性需要优先改善.基于这些发现,提出了有针对性的社区适老步行环境改善策略,旨在为未来的政策设计和养老服务提供参考和借鉴.
Population aging is a prominent phenomenon in many countries.It occurs as a result of declining fertility rates,increased life expectancy,and advancements in healthcare and living standards.The shift in age distribution presents significant challenges and opportunities for societies worldwide,impacting various aspects of social,economic,and healthcare systems."Actively addressing population aging"has recently been pursued as China's national strategy.Consequently,cities,especially those in China,must be motivated to adapt to an aging society,strive to satisfy the multifaceted needs of seniors,adopt a positive attitude and mindset,and undertake constructive actions while implementing pertinent policies to support the creation of age-friendly environments that enable seniors to live independently and actively participate in society. Walking,as a form of physical activity,offers many health benefits,including improved blood circulation,prevention of cardiovascular disease,cognitive decline mitigation,lowered risk of depression and dementia,boosted immunity,and enhanced cardiorespiratory function.Despite these benefits,walking has gradually been marginalized due to rapid urbanization and increased car usage,leading to a rise in obesity and cardiovascular diseases among urban dwellers and a decline in overall health.Therefore,it is imperative to reframe the significance of walking in urban life,advocating for increased walking and decreased reliance on cars.Furthermore,walking offers additional benefits for seniors.On one hand,it improves seniors'physical function,including balance and stability,which are crucial for maintaining independence and preventing falls,a common concern among seniors.This contributes significantly to"healthy aging",promoting longevity and overall well-being.On the other hand,walking encourages social interaction and community engagement,fostering connections with others and combating feelings of isolation and loneliness,a prevalent issue among seniors.These social interactions not only improve mental health but also contribute to a sense of belonging,ultimately enhancing quality of life and well-being.Thus,walking plays a pivotal role in supporting"active aging".Considering the multitude of advantages associated with walking,it becomes imperative to create a walking environment tailored to the needs of senior citizens,especially as the global population continues to age.In other words,as the demographic shift towards an aging population accelerates worldwide,ensuring that the walking environment accommodates the needs and preferences of seniors becomes increasingly important.This necessitates proactive measures by urban planners,policymakers,and community stakeholders. Satisfaction with the walking environment(or walking satisfaction),as a kind of transportation satisfaction,warrants attention in research aimed at enhancing the walking environment.At present,more research is needed on seniors'satisfaction with the walking environment.Moreover,conventional transportation satisfaction modeling approaches,including Pearson/Spearman correlation analysis,Logit model,ordered Logit/Probit model,and path analysis,often make rigid assumptions about the relationship between independent and dependent variables,neglecting potential nonlinearities and threshold effects.Therefore,the application of machine learning models becomes necessary for a more comprehensive analysis. In light of the above issues,this study conducts extensive questionnaire surveys among seniors in Chengdu,China,with a sample size of 1 012 respondents.A random forest model,a powerful ensemble learning algorithm,is developed to assess the relative importance of individual attributes of the walking environment in predicting overall walking satisfaction among seniors.Additionally,importance-performance analysis(IPA),a strategic management tool commonly used in marketing research to evaluate and prioritize various attributes or features of a product,service,or brand,is employed to prioritize the attributes of the walking environment from the seniors'perspective. The findings of this study are presented below.Random forest modeling results highlight that streetscape aesthetics,access to living amenities,walkway width,and seat availability are the attributes of the highest importance.Conversely,seniors attribute minimal value to intersection spacing and street connectivity.Moreover,IPA results emphasize the need to enhance seat availability,streetscape aesthetics,and access to recreational facilities and commercial centers.Lastly,targeted policy recommendations and strategic insights are proposed. This study serves as a valuable reference for guiding the enhancement of an age-friendly walking environment,a crucial consideration in urban planning and community development amidst the era of population aging.By examining various attributes and factors that contribute to the creation of an age-friendly walking environment,this study provides insights that can inform urban planners,architects,policymakers,and community stakeholders in their efforts to design and improve the walking environment,particularly within the context of population aging.
杨林川;唐祥龙;杨皓森;石亚灵
西南交通大学建筑学院成都理工大学地理与规划学院
土木建筑
人口老龄化随机森林重要性—绩效分析步行满意度老年友好步行环境老年友好城市
population agingrandom forestimportance-performance analysiswalking satisfactionage-friendly walking environmentage-friendly city
《西部人居环境学刊西部人居环境学刊》 2024 (002)
35-40 / 6
国家自然科学基金面上项目(52278080);北京大学—林肯中心年度研究基金(FS12-20211215-YLC)
评论