Increasing adoption rates at animal shelters: a two-phase approach to predict length of stay and optimal shelter allocation
Open accessThe DOI matches its Crossref record; 46 of the 47 references it cites were found in a registry or an index; 5 of the 14 sources its claims cite were checked for support.
Abstract
Among the 6-8 million animals that enter rescue shelters every year, nearly 3-4 million are euthanized, and 10-25% are put to death specifically because of shelter overcrowding each year.
Overview
CoChat Summary
## TL;DR
This study uses animal shelter intake/outcome data to predict animals’ length of stay with machine learning, aiming to reduce euthanization and improve adoption rates. Gradient boosting performed best overall, and the paper proposes a second-phase goal programming model to recommend shelter relocation decisions.
## Key findings
- The cleaned dataset contained 114,256 animal records, including 50,466 cat records and 63,790 dog records.
- Across the tested models, gradient boosting performed best overall, with the highest precision, recall, and F1 score; random forest was generally second best, and logistic regression performed worst.
- For overall average performance across metrics, the paper reports gradient boosting as the most proficient model at almost 60%; the performance gap was nearly 34% for precision, 39% for recall, and 32% for F1 score.
- Important predictors repeatedly identified included age (especially puppy, super senior, senior), size (large and small), and multicolor; gender, animal type, middle age, and medium-sized animals were not significant in their results.
## Methods
- The study used a two-phase design: phase 1 predicts length of stay with machine learning; phase 2 is a proposed goal programming approach for shelter relocation decisions.
- Data were assembled from public state/city shelter datasets and from shelters contacted across the U.S.; the paper says the final study focused on California, Kentucky, Colorado, Arizona, and Indiana and on dogs and cats.
- Four classification algorithms were compared on the cleaned data: logistic regression, artificial neural network, random forest, and gradient boosting; the dataset was split 80% training / 20% testing, with 10-fold cross-validation on training data and tuning via grid search for RF, GB, and ANN.
- Performance was evaluated using precision, recall (sensitivity), and F1 score.
## Limitations
- The paper states three limitations: lack of behavioral data, limited sample size, and use of simple algorithms.
- The text also makes evident that the data are not nationally comprehensive: there is no nationwide database, and only a subset of shelters responded to data requests, which likely limits representativeness.
- The paper notes that many shelters lacked the staff/resources to provide data, which constrained sample collection.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures









Verification
Claims the paper makes
2 Valid · 4 Not valid · 5 Not checked · 63 Own results and interpretations
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Finally, the gradient boosting and random forest algorithms perform well when predicting the very high length of stay at around 70–80%.”
§ Increasing adoption rates at animal shelters: a two-phase approach to predict length of stay and optimal shelter allocation > Results > Machine learning results
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“From Fig. 5, it is observed that the number of days a dog stays in the shelter decreases as the age increases.”
§ Increasing adoption rates at animal shelters: a two-phase approach to predict length of stay and optimal shelter allocation > Results > Exploratory Data results
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Specifically, we observe that older-aged animals (senior and/or super senior) appear as a significant factor for every algorithm.”
§ Increasing adoption rates at animal shelters: a two-phase approach to predict length of stay and optimal shelter allocation > Results > Machine learning results
hedgedstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Female cats that are returned have the highest variance for days spent in the shelter.”
§ Increasing adoption rates at animal shelters: a two-phase approach to predict length of stay and optimal shelter allocation > Results > Exploratory Data results
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“However, the performance of these models in predicting the medium length of stay for the given dataset is low for all three-performance metrics.”
§ Increasing adoption rates at animal shelters: a two-phase approach to predict length of stay and optimal shelter allocation > Results > Machine learning results
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
15 more claims
Create a free account to see all 74 claims and to verify the 2 references still open.
- Probably not a substantial claim. 36
- It may state two separate claims as one. 27
Citations
Cites
LeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
Machine learning: Trends, perspectives, and prospects
Jordan, M. I., & Mitchell, T. M. (2015) · Science · doi:10.1126/science.aaa8415 · Cited by 9,955 · cited 9×
“Xl≤ tl;y;s ∀l ð6Þ Equation [21] sets an upper limit on the length of stay category if the shelter l is assigned as the destination location.”
BackgroundInfluentialEnsemble methods in machine learning
Dietterich, T. G. (2000) · Lecture Notes in Computer Science · doi:10.1007/3-540-45014-9_1 · Cited by 8,408
Friedman, J., Hastie, T., & Tibshirani, R. (2000) · The Annals of Statistics · doi:10.1214/aos/1016218223 · Cited by 6,884
Random forests for classification in ecology
Cutler, D. R., Edwards, T. C., Beard, K. H., Cutler, A., Hess, K. T., Gibson, J., & Lawler, J. J. (2007) · Ecology · doi:10.1890/07-0539.1 · Cited by 4,950
Purposeful selection of variables in logistic regression
Bursac, Z., Gauss, C. H., Williams, D. K., & Hosmer, D. W. (2008) · Source Code for Biology and Medicine · doi:10.1186/1751-0473-3-17 · Cited by 4,035
Rokach, L. (2010) · Artificial Intelligence Review · doi:10.1007/s10462-009-9124-7 · Cited by 2,657 · cited 2×
“This process of re-weighing is done until a weighted vote of weak classifiers is combined into a robust classifier that determines the final labels or classification [46].”
BackgroundWaller, M. A., & Fawcett, S. E. (2013) · Journal of Business Logistics · doi:10.1111/jbl.12010 · Cited by 1,522
Machine learning and data mining methods in diabetes research
Kavakiotis, I., Tsave, O., Salifoglou, A., Maglaveras, N., Vlahavas, I., & Chouvarda, I. (2017) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2016.12.005 · Cited by 1,445
Data Mining: Concepts, Models, Methods, and Algorithms
(2003) · Technometrics · doi:10.1198/tech.2003.s785 · Cited by 1,369
Cited by
Predicting demand for air taxi urban aviation services using machine learning algorithms
Rajendran, S., Srinivas, S., & Grimshaw, T. (2021) · Journal of Air Transport Management · doi:10.1016/j.jairtraman.2021.102043 · Cited by 102
Self-citationRajendran, S., Chamundeswari, S., & Sinha, A. A. (2022) · Social Sciences & Humanities Open · doi:10.1016/j.ssaho.2022.100357 · Cited by 56
Self-citationAbandonment of dogs in Latin America: Strategies and ideas
Mota‐Rojas, D., Maldonado, N. A. C., Lezama-García, K., Sepiurka, L., & Garcia, R. D. C. M. (2021) · Veterinary World · doi:10.14202/vetworld.2021.2371-2379 · Cited by 82
Neuroimmunology of rabies: New insights into an ancient disease
Bastos, V. A. F., da Silva, V. P., Rodrigues, É. D. L., Moraes, C. N. D. S., Nóbile, A. L., Fonseca, D. L. M., Souza, K., do Vale, F. Y. N., Filgueiras, I. S., Schimke, L. F., et al. (2023) · Journal of Medical Virology · doi:10.1002/jmv.29042 · Cited by 45
Rajendran, S. (2021) · Expert Systems with Applications · doi:10.1016/j.eswa.2021.115056 · Cited by 32
Self-citation
Metadata
- Authors
- Bradley, J., & Rajendran, S.
- Affiliations
- University of Missouri (US)
- Corresponding author
- Rajendran, Suchithra
- ORCID
- Rajendran, Suchithra
- Venue
- BMC Veterinary Research
- Published
- 2021-02-05
- Volume
- 17
- Issue
- 1
- Pages
- 70-70
- ISSN
- 1746-6148 (ISSN-L)
- PMC
- PMC7863531
- OpenAlex
- W3127841828
- Semantic Scholar
- 9cb46124bdf7acf55e3a621352b93692ab4ac949
- Cited by
- 55
- References
- 47
- Citations by year
- 2026: 18 · 2025: 11 · 2024: 5 · 2023: 9 · 2022: 7 · 2021: 5
- FWCI
- 4.16
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by
- creativecommons.org/licenses/by/4.0/ (tdm)
- creativecommons.org/licenses/by/4.0/ (vor)
- Locations
- BMC Veterinary Research (publishedVersion, cc-by)
- Europe PMC (PubMed Central) (publishedVersion, other-oa)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Topics
- Physical Sciences › Computer Science › Information Systems › Data Mining and Machine Learning Applications
- Social Sciences › Law › Animal Law and Welfare
- Health Sciences › Health Professions › Health Information Management › Artificial Intelligence in Healthcare
- Keywords
- length of stay prediction, animal euthanasia, gradient boosting, random forest, artificial neural networks, logistic regression
- MeSH terms
- Machine Learning, Age Factors, Animals, Animal Welfare, Cats, Dogs, Female, Length of Stay, Male, Models, Statistical, Euthanasia, Animal
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Increasing adoption rates at animal shelters: a two-phase approach to predict length of stay and optimal shelter allocation. https://app.cochat.ai/papers/10.1186/s12917-020-02728-2
- The paper
- Bradley, J., & Rajendran, S. (2021). Increasing adoption rates at animal shelters: a two-phase approach to predict length of stay and optimal shelter allocation. BMC Veterinary Research. https://doi.org/10.1186/s12917-020-02728-2
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.