IPSO Parameter Settings for Rail Freight Volume Prediction, 2013-2022

114 monthly observations from the Lanzhou Freight Center in China, spanning 2013 to 2022. The dataset contains initial parameter settings for an Improved Particle Swarm Optimization (IPSO) algorithm, authored by Yunbo Gao and last updated in May 2026. It was used to train a hybrid FIG-IPSO-SVM model for predicting freight volume trends with uncertainty quantification.

Use Cases

• Benchmarking particle swarm optimization variants based on the described dynamic inertia weights and mutation operators.
• Training granular-level regression models for uncertainty quantification in time-series forecasting.
• Replicating the FIG-IPSO-SVM methodology for freight or logistics volume prediction.
• Analyzing the impact of parameter initialization on model convergence and prediction interval width.

Strengths

• Dataset is small (5.5 KB), facilitating quick download and inspection.
• Covers a specific 9-year time range (2013–2022) from a single freight center, providing temporal consistency.
• Parameters are associated with a model that achieved a reported mean maxima of absolute percentage error of 5.03% and a tight prediction interval.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset's scope is limited to parameter settings for a specific algorithm applied to a single case study.

Provenance

Source

Yunbo Gao via figshare.

Collection Method

Derived from 9-year operational records of the Lanzhou Freight Center.

Time Range

2013–2022

Freshness

Last updated 2026-05-07 17:39:43; freshness should be verified.

Geography

Lanzhou Freight Center, China