LSTM-CatBoost: Model Evaluation Results for ADSS Optical Cable Vibration Recognition

Evaluation results for a hybrid LSTM-CNN-CatBoost-GSSSA model designed to detect external force damage on underground All-Dielectric Self-Supporting (ADSS) optical cables. The dataset, shared by Xiaojuan Chen on figshare, contains performance metrics from experiments, including an accuracy of 97.6%. It was last updated on May 4, 2026.

Use Cases

• Benchmarking vibration signal recognition models based on the reported accuracy comparisons with SVM, CNN, and other hybrid models.
• Analyzing the computational efficiency of hybrid deep learning models for real-time monitoring applications.
• Optimizing hyperparameters for CatBoost models using evolutionary algorithms like the golden sine–enhanced Sparrow Search Algorithm (GSSSA).
• Studying feature fusion techniques for combining temporal, frequency, and spatial characteristics from sensor data.

Strengths

• Reported model accuracy of 97.6%, with specific performance gains over six baseline models detailed in the description.
• Dataset is openly licensed under CC-BY-4.0, facilitating reuse and redistribution.
• File size is 5.5 KB, indicating a compact and likely well-structured set of results.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for certain statistical analyses.
• The dataset's specific experimental setup and raw input data are not provided, limiting reproducibility of the training process.

Provenance

Source

figshare, author Xiaojuan Chen.

Collection Method

Likely contains experimental results from testing the proposed LSTM-CNN-CatBoost-GSSSA model on vibration signal data.

Time Range

null

Freshness

Last updated 2026-05-04 17:27:07; freshness should be verified.

Geography

null