Abstract
The change of number delivery goods through PT. Pos Regional VII East Java Indonesia indicates that the trend of increasing and decreasing the delivery of documents and non-documents in PT. Pos Regional VII East Java Indonesia is strongly influenced by conditions outside of PT. Pos Regional VII East Java Indonesia so that the prediction the number of document and non-documents requires a model that can accommodate it. Based on the time series plot monthly data fluctuations occur from 2013-2016 then the model is done using ARIMA or seasonal ARIMA and selected the best model based on the smallest AIC value. The results of data analysis about the number of shipments on each product sent through the Sub-Regional Postal Office VII East Java indicates that there are 5 post offices of 26 post offices entering the territory. The largest number of shipments is available on the PPB (Paket Pos Biasa is regular package shipment/non-document ) and SKH (Surat Kilat Khusus is Special Express Mail/document) products. The time series model generated is largely a Random walk model meaning that the number of shipment in the future is influenced by random effects that are difficult to predict. Some are AR and MA models, except for Express shipment products with Malang post office destination which has seasonal ARIMA model on lag 6 and 12. This means that the number of items in the following month is affected by the number of items in the previous 6 months.
Original language | English |
---|---|
Article number | 012056 |
Journal | Journal of Physics: Conference Series |
Volume | 974 |
Issue number | 1 |
DOIs | |
Publication status | Published - 22 Mar 2018 |
Event | 3rd International Conference on Mathematics: Pure, Applied and Computation, ICoMPAC 2017 - Surabaya, Indonesia Duration: 1 Nov 2017 → 1 Nov 2017 |
Keywords
- ARIMA
- documents
- non-documents
- post office
- shipments