Evaluating a survey of public livability perceptions and quality-of-life indicators: Considering freight-traffic impact

Abstract

The concept of livability is often used for integrating community quality of life, transportation facility access and neighborhood characteristics while supporting sustainability goals. Quality of life can be very difficult to measure because of the abstract nature of the concept and the varied factors that influence an individual's perceptions. Therefore, the current pilot-scale study examines whether an application of forecasting techniques using machine learning models on data from a previous stakeholder perception survey is possible and will aid in extracting quality of life patterns not only through selecting adequate prediction models but also by providing the most relevant subset of features (indicators) in each related model. This methodology has been used for similar problems in other domains, but has not been applied to livability. The results of this study provide evidence that there is a rule relating neighborhood perceptions to participants' livability scoring systems that can be revealed through machine learning techniques. The results are also consistent with a previous analytical hierarchy process (AHP) approach; however, the current methodology is able to uncover more apparent impact of freight on livability perceptions than was revealed through the previous AHP. Although the pilot results are promising, it is believed that with additional research, larger datasets, and data from multiple settings, more efficient livability indicators can be identified, adopted, and employed for planning purposes.

Publication Title

International Conference on Sustainable Infrastructure 2017: Methodology - Proceedings of the International Conference on Sustainable Infrastructure 2017

Link to Full Text

Share

COinS