This study investigates the use of a Long Short-Term Memory (LSTM) neural network model to forecast ferry passenger traffic at two ports in the Philippines - Batangas Port and Mindoro Port. The study uses monthly passenger traffic data from 2016 to 2022 obtained from the Philippine Ports Authority (PPA).
The key highlights and insights from the study are:
The proposed two-layer LSTM neural network model achieved an average forecasting accuracy of 72% for Batangas Port and 74% for Mindoro Port, as measured by the Mean Absolute Percentage Error (MAPE) metric.
The LSTM model was able to capture the temporal patterns in the ferry passenger traffic data and provide reasonable forecasts for the test sets.
The study recommends further investigation and comparison of the LSTM model with other forecasting techniques, as well as exploring ways to optimize the LSTM model's performance, such as testing different optimization techniques and combining it with other methods.
The ability to accurately forecast ferry passenger traffic can provide valuable insights for transportation and passenger management, enabling better decision-making and resource planning.
To Another Language
from source content
arxiv.org
Thông tin chi tiết chính được chắt lọc từ
by Daniel Fesal... lúc arxiv.org 05-06-2024
https://arxiv.org/pdf/2405.02098.pdfYêu cầu sâu hơn