TY - GEN
T1 - Self-Adaptive Particle Filter Based Time Series Prediction of Online Retailer Daily Sale
AU - Chen, Daoyuan
AU - Liu, Qinyi
AU - Yang, Xuanhao
AU - Yu, Xinyu
AU - Ma, Fei
AU - Su, Jionglong
N1 - Funding Information:
This study is supported by the National Natural Science Foundation of China (Grant No. 61501380), and Neusoft Corporation, item number SKLSAOP1702.
Publisher Copyright:
© 2019 IEEE.
PY - 2019/10
Y1 - 2019/10
N2 - Daily sale prediction is an essential step for industries to control inventory and reduce economic loss. In the past, it has been estimated using Kalman filter, with the restriction of linearity and Gaussian. In this study, a model combining parallel Sampling Importance Resampling Filter with an interacting multiple model is utilized to predict the future daily sale of products. The methodology is tested over a sale record of 80 products of a local online retailer in the past 400 days. The data of the last 30 days are used to verify the accuracy. Our experiments indicate that: 1) 29.33% of predicted values produced by the proposed method are within 10% fluctuation of true inventory data; (2) Better accuracy and efficacy can be achieved if either interacting multiple-model with different noise variances or parallel Sampling Importance Resampling Filter is applied; (3) Combination of the above two methods with suitable parameter settings may generate better performance, compared to the cases where only one of them is applied.
AB - Daily sale prediction is an essential step for industries to control inventory and reduce economic loss. In the past, it has been estimated using Kalman filter, with the restriction of linearity and Gaussian. In this study, a model combining parallel Sampling Importance Resampling Filter with an interacting multiple model is utilized to predict the future daily sale of products. The methodology is tested over a sale record of 80 products of a local online retailer in the past 400 days. The data of the last 30 days are used to verify the accuracy. Our experiments indicate that: 1) 29.33% of predicted values produced by the proposed method are within 10% fluctuation of true inventory data; (2) Better accuracy and efficacy can be achieved if either interacting multiple-model with different noise variances or parallel Sampling Importance Resampling Filter is applied; (3) Combination of the above two methods with suitable parameter settings may generate better performance, compared to the cases where only one of them is applied.
KW - Interacting multiple models
KW - nonstationary time series
KW - online retailer
KW - parallel Sampling Importance Resampling Filter
KW - real-time prediction
UR - http://www.scopus.com/inward/record.url?scp=85079175837&partnerID=8YFLogxK
U2 - 10.1109/CISP-BMEI48845.2019.8965758
DO - 10.1109/CISP-BMEI48845.2019.8965758
M3 - Conference Proceeding
AN - SCOPUS:85079175837
T3 - Proceedings - 2019 12th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics, CISP-BMEI 2019
BT - Proceedings - 2019 12th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics, CISP-BMEI 2019
A2 - Li, Qingli
A2 - Wang, Lipo
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 12th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics, CISP-BMEI 2019
Y2 - 19 October 2019 through 21 October 2019
ER -