An empirical analysis of intention of use for bike-sharing system in China through machine learning techniques

Abstract

Sharing bicycles, as boosted by the advanced mobile technologies, is expected to mitigate the traffic congestion and air pollution issues in China. A survey study was conducted with 335 valid samples to identify the key factors that influence the customers' intention of use for bike-sharing system and quantify the corre-sponding importance. Five machine learning techniques for classi-fication are applied and results are compared. The best performed technique is selected to prioritise and quantify the importance level of the influencing factors. The results indicate that the perceived ease of use is the most significant factor for the intention to use sharing bikes.

What drives e-hailing apps adoption? An analysis of behavioral factors through fuzzy AHP

“…The finding is aligned to the study of Mallat et al (2008) and Schmitz et al (2016), which indicates a significant impact of perceived ease of use on the acceptance of mobile technology in public transportation service. Zhou et al (2020) in their study found perceived ease of use as the most significant contributor to the intention to use a bike-sharing system. However, the hypothesis regarding the positive effect of perceived ease of use on user satisfaction with e-hailing apps was not supported in a study conducted by Joia and Altieri (2017, 2018).…”

“…In the context of e-hailing apps, not many studies have been conducted so far that have explored factors affecting user adoption of these apps. The studies the researchers found in this context have used different theoretical models instituting various additional variables in those such as: subjective norms, perceived usefulness, perceived ease of use, compatibility, trust, complexity, relative advantage (Joia and Altieri, 2018); perceived usefulness, perceived ease of use, perceived privacy risk, trust (Zhang et al , 2017); perceived usefulness, perceived ease of use, subjective norm, perceived behavior control (Giang et al , 2017); performance expectancy, effort expectancy, social influence, facilitating conditions (Haba and Dastane, 2018); performance expectancy, effort expectancy, social influence, perceived risk, perceived cost, facilitating conditions (Jiang and Zhu, 2018); perceived usefulness, subjective norms, perceived risk, perceived playfulness, perceived price level (Lim et al , 2018); relative advantages, perceived ease of use, trialability, social influence, physical security (Ruangkanjanases and Techapoolphol, 2018); perceived usefulness, subjective norms, perceived ease of use, perceived risk (Weng et al , 2017); subjective norms, perceived usefulness, perceived ease of use, compatibility, relative advantage, safety (Arumugama et al , 2020); perceived usefulness, perceived ease of use, subjective norm, perceived behavioral control (Haldar and Goel, 2019); performance expectancy, effort expectancy, trust, enjoyment (Razi et al , 2019); perceived usefulness, perceived ease of use (Suhud et al , 2019); perceived usefulness, subjective norm, perceived risk, perceived playfulness, perceived price level, perceived ease of use, compatibility (Peng et al , 2014); perceived usefulness, perceived ease of use, social influence, trust, perceived risk (Zhang et al , 2016); perceived usefulness, perceived usability, perceived trust (Karim et al , 2020); perceived usefulness, perceived ease of use, perceived discount, financial risk, security risk, privacy risk (Zhou et al , 2020). It has been observed that the factors that have been used most frequently under these studies are: perceived usefulness, perceived ease of use, compatibility, perceived risk, trust, subjective norms, social influence, facilitating conditions, relative advantage and perceived behavioral control.…”

A multi-method analytical approach to predicting young adults’ intention to invest in mHealth during the COVID-19 pandemic

“…A five-fold cross-validation technique was applied to avoid overfitting. Additionally, early stopping criteria were employed in ANN to halt the training process after a large number of epochs where the loss has not been improved ( Zhou et al, 2021 ).…”

“…Numerous techniques for SA have been described in scientific literature. In recent literature ( Zhou et al, 2021 ), Sobol’s indices ( Sobol, 2001 ) have gained prominence because of variance-based methods for substitute model concepts. Sobol’s SA can determine the contribution of each predictor variable and their interconnections to the overall model output variance ( Zhang et al, 2015 ).…”