Dataset for a Multimodal Machine Learning System for Non-Invasive Detection of Varroa Destructor Infestations in Honey Bee Colonies
Fuente:
Zenodo
Enregistré dans:
| Auteur principal: | |
|---|---|
| Format: | Recurso digital |
| Publié: |
Zenodo
2026
|
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866901085047750656 |
|---|---|
| author | Gaba, Parth |
| author_facet | Gaba, Parth |
| contents | <p>This dataset contains processed field deployment data associated with a multimodal machine learning system for non-invasive detection of Varroa destructor infestations in Apis mellifera (honey bee) colonies.</p> <p>The dataset includes three real-world deployments representing distinct colony health conditions:</p> <ul> <li>A 30-day continuous deployment on a confirmed healthy colony (0 mites per 100 bees), used to validate baseline system behavior and false positive rates.</li> <li>A 30-day deployment on a mildly infested colony (2–4 mites per 100 bees), capturing early-stage distributional shifts without threshold-triggered alerts.</li> <li>A 1-day deployment on a clearly infested colony (8+ mites per 100 bees), demonstrating threshold exceedance and alert generation under high infestation conditions.</li> </ul> <p>Each dataset consists of timestamped fused probability scores generated from a confidence-weighted fusion of multiple sensing modalities, along with corresponding alert flags and hive status labels. The fused score represents the system’s estimate of infestation likelihood at each observation interval.</p> <p>Due to the large size of raw multimodal recordings (~300 GB), this release provides a processed and downsampled dataset suitable for reproducing the analysis and results reported in the associated study.</p> <p>This dataset enables evaluation of system performance across varying infestation levels and provides evidence of real-world deployment behavior under continuous monitoring conditions.</p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_19293834 |
| institution | Zenodo |
| language | |
| publishDate | 2026 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | Dataset for a Multimodal Machine Learning System for Non-Invasive Detection of Varroa Destructor Infestations in Honey Bee Colonies Gaba, Parth <p>This dataset contains processed field deployment data associated with a multimodal machine learning system for non-invasive detection of Varroa destructor infestations in Apis mellifera (honey bee) colonies.</p> <p>The dataset includes three real-world deployments representing distinct colony health conditions:</p> <ul> <li>A 30-day continuous deployment on a confirmed healthy colony (0 mites per 100 bees), used to validate baseline system behavior and false positive rates.</li> <li>A 30-day deployment on a mildly infested colony (2–4 mites per 100 bees), capturing early-stage distributional shifts without threshold-triggered alerts.</li> <li>A 1-day deployment on a clearly infested colony (8+ mites per 100 bees), demonstrating threshold exceedance and alert generation under high infestation conditions.</li> </ul> <p>Each dataset consists of timestamped fused probability scores generated from a confidence-weighted fusion of multiple sensing modalities, along with corresponding alert flags and hive status labels. The fused score represents the system’s estimate of infestation likelihood at each observation interval.</p> <p>Due to the large size of raw multimodal recordings (~300 GB), this release provides a processed and downsampled dataset suitable for reproducing the analysis and results reported in the associated study.</p> <p>This dataset enables evaluation of system performance across varying infestation levels and provides evidence of real-world deployment behavior under continuous monitoring conditions.</p> |
| title | Dataset for a Multimodal Machine Learning System for Non-Invasive Detection of Varroa Destructor Infestations in Honey Bee Colonies |
| url | https://doi.org/10.5281/zenodo.19293834 |