<![CDATA[The Jakarta government has used digital infrastructure, such as online platforms and software applications, to implement these elements. However, there is still room for improvement in maximizing benefits for the city and its residents. One area that needs to be optimized is the development of smart mobility prediction models and improving the performance of existing models. In this study, Gaussian hidden markov model (GMM) and autoregressive integrated moving average (ARIMA) algorithms were used for predictable mobility monitoring to decipher congestion in Jakarta. The parameters used in detecting stay points are a time threshold of 20 minutes, and a distance threshold of 200 meters. The evaluation results showed that the ARIMA model test obtained a root mean square error (RMSE) value of 162,766, showing a fairly high error. The evaluation results for prediction using GHMM predicting mobility to support the Jakarta Smart City program on the test data were calculated using the accuracy model and RMSE model. The performance of GHMM gets an accuracy of 76.90% and RMSE of 1,641. The evaluation value of GHMM can be said to be good enough to model mobility data.]]>
Copyrights © 2023
