Sensitivity Analysis of Hyperparameters for Hyperspectral Band Selection

5.5 KB of tabular data from figshare, authored by Dandan He and last updated on 2026-04 17. The dataset likely contains results from a sensitivity analysis of an improved band selection algorithm for hyperspectral images. The proposed SSGIE-KFCM method reportedly achieved over 90% average classification accuracy on the Indian Pines and Pavia University datasets.

Use Cases

• Benchmarking band selection algorithms based on reported classification accuracy and AUC metrics.
• Analyzing the computational efficiency of optimization methods based on reported runtime reductions.
• Studying the impact of hyperparameter adjustments on algorithm convergence and performance.
• Developing lightweight solutions for real-time remote sensing processing based on the described engineering applications.

Strengths

• Results are validated on two standard hyperspectral datasets (Indian Pines and Pavia University).
• Reported performance metrics include specific numbers: over 90% average accuracy, 0.958 AUC, and 40% of traditional method runtime.
• Algorithm performance is detailed with specific computation times (0.058s and 0.172s).

Limitations

• Row count and column-level documentation are absent; field semantics must be inferred after download.
• The dataset is very small (5.5 KB), indicating limited scope, likely containing summary results rather than raw data.

Provenance

Source

figshare, author Dandan He.

Collection Method

Likely contains experimental results from a research study on an improved band selection algorithm.

Time Range

null

Freshness

Last updated 2026-04-17 17:37:45; freshness should be verified.

Geography

null