1. Zhejiang Hongcheng Computer Systems Company Limited, Hangzhou 310009, China 2. College of Computer Science and Technology, Zhejiang University, Hangzhou 310027, China
The proposed prediction model adopted a time series segmentation strategy. Firstly, each time series was split into segments and convolved separately by using convolutional neural network (CNN). Then, gated recurrent unit (GRU) was utilized to mine the time series information. Finally, relevant context information (weather conditions and holidays) of the prediction time was added to predict short-term tourist number. Experiments on the gate dataset and vehicle monitoring dataset of a scenic area show that multi-factor perceived short-term tourist number prediction model based on GRU considers multi context factors and the differences between multi time series. The root mean square error (RMSE) and mean absolute percent error (MAPE) of the proposed model are smaller than that of the traditional models, which reveals that the model can effectively reduce short-term tourist number prediction error.
Jing-chang WANG,Ling CHEN,Shan-shan YU,Chen-shu JIANG,Yong WU. Multi-factor perceived short-term tourist number prediction model based on gated recurrent unit. Journal of ZheJiang University (Engineering Science), 2019, 53(12): 2357-2364.
Fig.1Architecture of multi-factor perceived short-term tourist number prediction model based on gated recurrent unit (MFPPM)
数据集
时间序列
参数符号
景区闸机 数据
景区内游客人数
xtot
进入景区游客人数
xin
离开景区游客人数
xout
监控点车辆 数据
本地客车进入数
xttbin
本地客车离开数
xttbout
本地轿车进入数
xttcin
本地轿车离开数
xttcout
外地本省客车进入数
xtfbin
外地本省客车离开数
xtfbout
外地本省轿车进入数
xtfcin
外地本省轿车离开数
xtfcout
外省客车进入数
xffbin
外省客车离开数
xffbout
外省轿车进入数
xffcin
外省轿车离开数
xffcout
Tab.1Time series obtained from gate dataset and vehicle monitoring dataset of scenic area
Fig.2CNN structure for time series segmentation strategy
Fig.3Diagram of gated recurrent unit (GRU)
网络层
超参
卷积层
Filters=3；Kernel_size=5；Strides=1；Activation=Relu
池化层
第1个池化层：Pool_size=3；第2个池化层：Pool_size=2
GRU
Units=d
全连接层
Units=1；Activation=LeakyRelu（alpha=0.3）
Tab.2Hype-rparameters setting of MFPPM
评估指标
RMSE
MAPE
m=420
193.63
150.16
m=480
190.98
121.08
m=540
193.92
123.65
m=600
197.67
132.82
Tab.3Effect of parameter m on results in MFPPM
评估指标
RMSE
MAPE
评估指标
RMSE
MAPE
d=50
195.73
136.54
d=70
190.98
121.08
d=60
192.93
123.56
d=80
197.35
129.70
Tab.4Effect of output dimension on results in MFAPM
模型
RMSE
MAPE
MFPPM-W/O-Weather
193.02
129. 96
MFPPM-W/O-Holiday
203.04
140.28
MFPPM-W/O-Vehicle
217.04
159.77
MFPPM
190.98
121.08
Tab.5Effect of different context factors on prediction results of short-term tourist numbers
模型
RMSE
MAPE
SVR-multi
272.63
147.19
GRU-multi
268.76
141.16
CNNGRU-multi
250.49
139.02
MFPPM-only
225.05
130.67
MFPPM
190.98
121.08
Tab.6Comparison of prediction effects by different short-term tourist number prediction models
[1]
CHU F L Forecasting tourism demand with ARMA-based methods[J]. Tourism Management, 2009, 30 (5): 740- 751
doi: 10.1016/j.tourman.2008.10.016
[2]
GUSTAVSSON P, NORDSTROM J The impact of seasonal unit roots and vector ARMA modelling on forecasting monthly tourism flows[J]. Tourism Economics, 2001, 7 (2): 117- 133
doi: 10.5367/000000001101297766
[3]
BOX G E P, PIERCE D A Distribution of residual autocorrelations in autoregressive-integrated moving average time series models[J]. Journal of the American Statistical Association, 1970, 65 (332): 1509- 1526
doi: 10.1080/01621459.1970.10481180
[4]
AHMED M S, COOK A R. Analysis of freeway traffic time-series data by using Box-Jenkins techniques [M]. [S. l.]: Transportation Research Board, 1979.
[5]
LIM C, MCALEER M Time series forecasts of international travel demand for Australia[J]. Tourism Management, 2002, 23 (4): 389- 396
doi: 10.1016/S0261-5177(01)00098-X
[6]
PAI P F, HONG W C An improved neural network model in forecasting arrivals[J]. Annals of Tourism Research, 2005, 32 (4): 1138- 1141
doi: 10.1016/j.annals.2005.01.002
[7]
JANACEK G Time series analysis forecasting and control[J]. Journal of Time Series Analysis, 2010, 31 (4): 303
[8]
WU C H, HO J M, LEE D T Travel-time prediction with support vector regression[J]. IEEE Transactions on Intelligent Transportation Systems, 2004, 5 (4): 276- 281
doi: 10.1109/TITS.2004.837813
[9]
HONG W C Traffic flow forecasting by seasonal SVR with chaotic simulated annealing algorithm[J]. Neurocomputing, 2011, 74 (12-13): 2096- 2107
doi: 10.1016/j.neucom.2010.12.032
[10]
CHEN K Y, WANG C H Using genetic algorithms to optimize support vector regression in tourism demand forecasting[J]. Journal of Outdoor Recreat Study, 2005, 18 (1): 47- 72
[11]
PAI P F, HONG W C, CHANG P T, et al The application of support vector machines to forecast tourist arrivals in Barbados: an empirical study[J]. International Journal of Management, 2006, 23 (2): 375
[12]
CHEN K Y, WANG C H Support vector regression with genetic algorithms in forecasting tourism demand[J]. Tourism Management, 2007, 28 (1): 215- 226
doi: 10.1016/j.tourman.2005.12.018
[13]
CHEN K Y Combining linear and nonlinear model in forecasting tourism demand[J]. Expert Systems with Applications, 2011, 38 (8): 10368- 10376
doi: 10.1016/j.eswa.2011.02.049
[14]
CHEN R, LIANG C Y, HONG W C, et al Forecasting holiday daily tourist flow based on seasonal support vector regression with adaptive genetic algorithm[J]. Applied Soft Computing, 2015, 26: 435- 443
doi: 10.1016/j.asoc.2014.10.022
CHO K, MERRIENBOER B V, GULCEHRE C, et al. Learning phrase representations using RNN encoder-decoder for statistical machine translation [J]. arXiv preprint. arXiv: 1406.1078, 2014.
[17]
LI Y F, CAO H. Prediction for tourism flow based on LSTM neural network [C] // International Conference on Identification, Information and Knowledge in the Internet of Things. Qufu: ELSEVIER, 2017: 277-287.
[18]
FU R, ZHANG Z, LI L. Using LSTM and GRU neural network methods for traffic flow prediction [C] // 2016 31st Youth Academic Annual Conference of Chinese Association of Automation. Wuhan: IEEE, 2016: 324-328.
[19]
GUO J, WANG Z, CHEN H. On-line multi-step prediction of short term traffic flow based on GRU neural network [C] // Proceedings of the 2nd International Conference on Intelligent Information Processing. Bangkok: ACM, 2017: 11.
[20]
王文庆, 郭恒, 范启富. 基于CNN与GRU的航空发动机剩余寿命预测 [C] // 第37届中国控制会议. 武汉: 中国学术期刊电子出版社, 2018: 546-551. WANG Wen-qing, GUO Heng, FAN Qi-fu. Predict remaining useful life of aerospace engine based on CNN and GRU [C] // Proceedings of the 37th Chinese Control Conference. Wuhan: China Academic Journal Electronic Publishing House, 2018: 546-551.
