| Summary: | 碩士 === 輔仁大學 === 電機工程學系 === 100 === Our design, which uses an adaptive calibration method or machine learning method to improve the accuracy of the taxi fare estimation, depends on a mashup of both electronic map application programming interface and global positioning system signals. Our design also combines with short message service (SMS), social networks and cloud storage to develop an application program on the smart phone with multifunction of real-time route planning, fare estimation and a security tracking system. The fare estimation is re-predicted automatically by using the calibration method. We divide the fare and delay of timing into several ranks, which, by using the calibration method, compensate the error of estimation and improve the accuracy of the taxi fare estimation program. Each time the program can record the fare and the timing delay. These data are converted to metadata and compensate the error of estimation at the next time. Reducing the cumulative distance increases the accuracy of the actual distance. The two-pronged approach with these two methods in prediction and simulation improves the accuracy of the taxi fare estimation program.
|
|---|
