Applied of feed-forward neural network and facebook prophet model for train passengers forecasting

Train is public transportation that is widely used by people in Indonesia. Due to the high level of comfort at a low price relatively. Based on released data by PT KAI, the number of train passengers has increased in almost every holiday season, thereby, it is suspected that there are seasonal patterns with fixed and random periods. In 2013, the government issued a policy related to infrastructure development, then, it caused the number of passengers significantly increasing in the following year. Thus, we need a model that can accommodate these patterns to forecast the number of train passengers accurately. The use of neural network methods such as Feed Forward Neural Network (FFNN), nowadays, becomes popular in facing big data including unexpected fluctuation on the data. Additionally, recently, Facebook announced an accurate method of forecasting, called Prophet model, for data which have trend, seasonality, holidays, and missing data. Hence, the forecast for monthly train passengers on this research is modelled by FFNN and Facebook Prophet. The result shows that Prophet performs better than FFNN. However, the difference in the value of MAPE is not too large.

[1]  Benjamin Letham,et al.  Forecasting at Scale , 2018 .

[2]  George Athanasopoulos,et al.  Forecasting: principles and practice , 2013 .

[3]  A. Harvey,et al.  Estimation Procedures for Structural Time Series Models , 1990 .

[4]  Yung-Cheng Lai,et al.  Enhanced Parametric Railway Capacity Evaluation Tool , 2009 .

[5]  W. Pitts,et al.  A Logical Calculus of the Ideas Immanent in Nervous Activity (1943) , 2021, Ideas That Created the Future.

[6]  Hao Li,et al.  Content preserving image translation with texture co-occurrence and spatial self-similarity for texture debiasing and domain adaptation. , 2021, Neural networks : the official journal of the International Neural Network Society.

[7]  Keith W. Hipel,et al.  Forecasting nonlinear time series with feed-forward neural networks: a case study of Canadian lynx data , 2005 .

[8]  Zaher Mundher Yaseen,et al.  RBFNN versus FFNN for daily river flow forecasting at Johor River, Malaysia , 2015, Neural Computing and Applications.

[9]  . Anamika,et al.  Electricity Price Forecasting and Classification Through Wavelet–Dynamic Weighted PSO–FFNN Approach , 2018, IEEE Systems Journal.

[10]  Frank Rosenblatt,et al.  PRINCIPLES OF NEURODYNAMICS. PERCEPTRONS AND THE THEORY OF BRAIN MECHANISMS , 1963 .

[11]  Mariano Gallo,et al.  Artificial Neural Networks for Forecasting Passenger Flows on Metro Lines , 2019, Sensors.

[12]  Christopher P. L. Barkan,et al.  Impact of Train Type Heterogeneity on Single-Track Railway Capacity , 2009 .