Relationship between Cab prices, Traffic flow and Various Cab Indicators, Location and Weather

Abstract

Cab price analysis is based on cab prices under different conditions and is widely used in navigation platforms, trip selection, and inter-company strategic competition. Multiple linear regression models are a commonly used method of price analysis that identifies the correlation between single or multiple factors and prices and continuously optimizes accuracy based on the analysis of variance. However, there is still a gap in research regarding the relationship of multiple factors on specific prices. To eliminate this limitation, this paper uses a multiple linear regression model to analyze the factors influencing cab prices and improve refinement. First, the dataset is preprocessed to deal with missing data, and categorical data, and remove useless data. Then, we construct linear regression models to perform correlation analysis by verifying the combination of a single weather attribute and a price attribute. To integrate the multi-attribute overlay, a composite multi-weather attribute combination is used for the analysis. To verify the validity of the analysis and forecast results, we perform significant objective tests, and goodness-of-fit tests, and calculate and compare parameters including mean absolute error, mean square error, root mean square and mean absolute percentage error. We compare the results of single and composite analyses. Where the accuracy of a composite one is better than that of a single one. The numerical experimental results verify the effectiveness of our method.