Regression Analysis Using Machine Learning Approaches for Predicting Container Shipping Rates
- Publisher:
- Springer Nature
- Publication Type:
- Conference Proceeding
- Citation:
- Lecture Notes in Networks and Systems, 2022, 450 LNNS, pp. 269-280
- Issue Date:
- 2022-01-01
Closed Access
| Filename | Description | Size | |||
|---|---|---|---|---|---|
| 978-3-030-99587-4_23.pdf | Published version | 1.29 MB |
Copyright Clearance Process
- Recently Added
- In Progress
- Closed Access
This item is closed access and not available.
Container freight rate forecasts are used by major stakeholders in the maritime industry, such as shipping lines, consumers, shippers, and others, to make operational decisions. Because container shipping lacks a structured forwards market, it must rely on forecasts for hedging reasons. This research is dedicated to investigating and predicting shipping containerised freight rates using machine learning approaches and real-time data to uncover superior forecasting methods. Ensemble models including Random Forest (RF) and Extreme Gradient Boosting (XGBoost), and deep learning, in particular Multi-Layer Perceptions (MLP) have all been used to provide data-driven predictions after initial feature engineering. These three regression-based machine learning (ML) models are used to predict the container shipping rates in the North American TransBorder Freight dataset from 2006 to 2021. It has been found that MLP surpasses ensemble models with a test accuracy rate of 97%. Although our findings are drawn from American shipping data, the proposed approach serves as a general method for other international markets.
Please use this identifier to cite or link to this item: