Innovative Machine Learning Approaches for Accurate and Ethical Depression Diagnosis: Insights and Recommendations

Abiodun Abdulghani Akanbi

doi:doi:10.11648/j.mlr.20251002.17

Research Article |

| Peer-Reviewed

Innovative Machine Learning Approaches for Accurate and Ethical Depression Diagnosis: Insights and Recommendations

Abiodun Abdulghani Akanbi^*

Published in Machine Learning Research (Volume 10, Issue 2)

Received: 10 July 2025 Accepted: 22 September 2025 Published: 19 December 2025

Views: Downloads:

Download PDF

Share This Article

Twitter
Linked In
Facebook

Abstract

This study advances mental-health diagnostics by integrating supervised and unsupervised machine-learning methods with a strong ethical lens. We review and contextualize the current literature on ML applications for depression detection, noting the heavy reliance of supervised models on labelled data and the comparative under-exploration of unsupervised approaches. Using the public Depression Dataset-which comprises actigraph recordings from depressed patients and healthy controls and includes demographic and clinical attributes such as timestamps, activity counts, gender, age and Montgomery-Åsberg Depression Rating Scale HYPERLINK "http://kaggle.com" \h score we preprocess and engineer features capturing circadian rhythms and variability in motor activity. We then apply multiple clustering algorithms (K-means, hierarchical clustering and DBSCAN) to identify latent subgroups of depression severity, evaluating cluster validity via the silhouette score, Davies-Bouldin index and Calinski-Harabasz index. A supervised support-vector machine classifier trained on labelled severity categories serves as a baseline, and we find that unsupervised clustering achieves competitive performance while revealing nuanced patterns not captured by labelled categories. We visualise cluster structures and compare performance metrics to illustrate the benefits and limitations of each method. Finally, we analyse cluster composition relative to gender and socio-economic variables to highlight potential biases in the data and underscore the need for fairness-aware models and interpretability. By combining supervised and unsupervised techniques and explicitly addressing ethical considerations, this work contributes to more accurate, transparent and equitable ML-based depression diagnosis.

Published in	Machine Learning Research (Volume 10, Issue 2)
DOI	10.11648/j.mlr.20251002.17
Page(s)	158-164
Creative Commons	This is an Open Access article, distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution and reproduction in any medium or format, provided the original work is properly cited.
Copyright	Copyright © The Author(s), 2025. Published by Science Publishing Group

Keywords

Machine Learning, Unsupervised Learning, Depression Diagnosis, Clustering, Ethical AI

References

[1]	O. Oyebode, F. Alqahtani, and R. Orji, "Using Machine Learning and Thematic Analysis Methods to Evaluate Mental Health Apps Based on User Reviews," IEEE Access, vol. 8, pp. 111141-111158, 2020. https://doi.org/10.1109/ACCESS.2020.3002176
[2]	C. Su, Z. Xu, J. Pathak, and F. Wang, "Deep learning in mental health outcome research: a scoping review," Translational Psychiatry, vol. 10, no. 1, p. 116, 2020. https://doi.org/10.1038/s41398-020-0801-z
[3]	A. M. Chekroud, J. Bondar, J. Delgadillo, G. Doherty, A. Wasil, M. Fokkema, Z. Cohen, D. Belgrave, R. DeRubeis, R. Iniesta, D. Dwyer, and K. Choi, "The promise of machine learning in predicting treatment outcomes in psychiatry," World Psychiatry, vol. 20, no. 2, pp. 154-161, Jun. 2021. https://doi.org/10.1002/wps.20882
[4]	T. Jiang, J. L. Gradus, and A. J. Rosellini, "Supervised Machine Learning: A Brief Primer," Behavior Therapy, vol. 51, no. 5, pp. 801-819, Oct. 2020. https://doi.org/10.1016/j.beth.2020.05.002
[5]	J. Sun, Q.-X. Dong, S.-W. Wang, Y.-B. Zheng, X.-X. Liu, T.-S. Lu, K. Yuan, J. Shi, B. Hu, L. Lu, and Y. Han, "Artificial intelligence in psychiatry research, diagnosis, and therapy," Asian Journal of Psychiatry, vol. 78, p. 103705, Dec. 2023. https://doi.org/10.1016/j.ajp.2023.103705
[6]	F. Li, H. Sun, B. B. Biswal, J. A. Sweeney, and Q. Gong, "Artificial intelligence applications in psychoradiology," Psychoradiology, vol. 1, no. 2, pp. 94-107, Jun. 2021. https://doi.org/10.1093/psyrad/kkab009
[7]	C. Blease, C. Locher, and M. Doraiswamy, "Artificial intelligence and the future of psychiatry: Qualitative findings from a global physician survey," Digital Health, vol. 6, p. 2055207620968355, Dec. 2020. https://doi.org/10.1177/2055207620968355
[8]	S. Bhadra and C. J. Kumar, "An insight into the diagnosis of depression using machine learning techniques: a systematic review," Journal of Medical Engineering & Technology, vol. 46, no. 8, pp. 749-771, Nov. 2022. https://doi.org/10.1080/03007995.2022.2038487
[9]	J. Benoit, H. Onyeaka, M. Keshavan, and J. Torous, "Systematic Review of Digital Phenotyping and Machine Learning in Psychosis Spectrum Illnesses," Harvard Review of Psychiatry, vol. 28, no. 5, pp. 296-304, Sep. 2020. https://doi.org/10.1097/HRP.0000000000000268
[10]	N. S. M. Shafiee and S. Mutalib, "Prediction of Mental Health Problems among Higher Education Student Using Machine Learning," I. J. Education and Management Engineering, vol. 6, pp. 1-9, Dec. 2020. https://doi.org/10.5815/ijeme.2020.06.01
[11]	Y. Wu, M. Daoudi, A. Amad, L. Sparrow, and F. D'Hondt, "Unsupervised Learning Method for Exploring Students' Mental Stress in Medical Simulation Training," in ICMI '20 Companion: Companion Publication of the 2020 International Conference on Multimodal Interaction, 2020, pp. 165-170. https://doi.org/10.1145/3395035.3425191
[12]	A. Le Glaz, Y. Haralambous, D. H. Kim-Dufor, P. Lenca, R. Billot, T. C. Ryan, J. Marsh, J. DeVylder, M. Walter, S. Berrouiguet, and C. Lemey, "Machine Learning and Natural Language Processing in Mental Health: Systematic Review," Journal of Medical Internet Research, vol. 23, no. 5, p. e15708, May 2021. https://doi.org/10.2196/15708
[13]	N. Alexander, D. C. Alexander, F. Barkhof, and S. Denaxas, "Using Unsupervised Learning to Identify Clinical Subtypes of Alzheimer's Disease in Electronic Health Records," Studies in Health Technology and Informatics, vol. 270, pp. 499-503, Oct. 2020. https://doi.org/10.3233/SHTI200210
[14]	H. Pelin, M. Ising, F. Stein, S. Meinert, T. Meller, K. Brosch,. and T. F. M. Andlauer, "Identification of transdiagnostic psychiatric disorder subtypes using unsupervised learning," Neuropsychopharmacology, vol. 46, pp. 1895-1905, Aug. 2021. https://doi.org/10.1038/s41386-021-01048-6
[15]	I. Straw and C. Callison-Burch, "Artificial Intelligence in mental health and the biases of language-based models," PLOS ONE, vol. 15, no. 12, p. e0240376, Dec. 2020. https://doi.org/10.1371/journal.pone.0240376
[16]	B. Kung, M. Chiang, G. Perera, M. Pritchard, and R. Stewart, "Unsupervised Machine Learning to Identify Depressive Subtypes," Healthcare Informatics Research, vol. 28, no. 3, pp. 256-266, Sep. 2022. https://doi.org/10.4258/hir.2022.28.3.256
[17]	I. Twamley, M. Dempsey, and N. Keane, "An Open Dialogue-informed approach to mental health service delivery: experiences of service users and support networks," Psychiatric Rehabilitation Journal, vol. 43, no 4.
[18]	Aafjes-van Doorn, K., Kamsteeg, C., Bate, J., & Aafjes, M. (2020). A scoping review of machine learning in psychotherapy research. Frontiers in Psychiatry, 11, 738466. https://doi.org/10.1080/10503307.2020.1808729
[19]	Ngumimi, K. I., Kim, S.-H., Jhon, M., Yang, H.-J., & Pant, S. (2023). A Review of Machine Learning and Deep Learning Approaches on Mental Health Diagnosis. Healthcare, 11(3), 285. https://doi.org/10.3390/healthcare11030285

Cite This Article

Plain Text BibTeX RIS

APA Style

Akanbi, A. A. (2025). Innovative Machine Learning Approaches for Accurate and Ethical Depression Diagnosis: Insights and Recommendations. Machine Learning Research, 10(2), 158-164. https://doi.org/10.11648/j.mlr.20251002.17

Copy | Download

ACS Style

Akanbi, A. A. Innovative Machine Learning Approaches for Accurate and Ethical Depression Diagnosis: Insights and Recommendations. Mach. Learn. Res. 2025, 10(2), 158-164. doi: 10.11648/j.mlr.20251002.17

Copy | Download

AMA Style

Akanbi AA. Innovative Machine Learning Approaches for Accurate and Ethical Depression Diagnosis: Insights and Recommendations. Mach Learn Res. 2025;10(2):158-164. doi: 10.11648/j.mlr.20251002.17

Copy | Download

@article{10.11648/j.mlr.20251002.17,
  author = {Abiodun Abdulghani Akanbi},
  title = {Innovative Machine Learning Approaches for Accurate and Ethical Depression Diagnosis: Insights and Recommendations},
  journal = {Machine Learning Research},
  volume = {10},
  number = {2},
  pages = {158-164},
  doi = {10.11648/j.mlr.20251002.17},
  url = {https://doi.org/10.11648/j.mlr.20251002.17},
  eprint = {https://article.sciencepublishinggroup.com/pdf/10.11648.j.mlr.20251002.17},
  abstract = {This study advances mental-health diagnostics by integrating supervised and unsupervised machine-learning methods with a strong ethical lens. We review and contextualize the current literature on ML applications for depression detection, noting the heavy reliance of supervised models on labelled data and the comparative under-exploration of unsupervised approaches. Using the public Depression Dataset-which comprises actigraph recordings from depressed patients and healthy controls and includes demographic and clinical attributes such as timestamps, activity counts, gender, age and Montgomery-Åsberg Depression Rating Scale  HYPERLINK "http://kaggle.com"  \h score we preprocess and engineer features capturing circadian rhythms and variability in motor activity. We then apply multiple clustering algorithms (K-means, hierarchical clustering and DBSCAN) to identify latent subgroups of depression severity, evaluating cluster validity via the silhouette score, Davies-Bouldin index and Calinski-Harabasz index. A supervised support-vector machine classifier trained on labelled severity categories serves as a baseline, and we find that unsupervised clustering achieves competitive performance while revealing nuanced patterns not captured by labelled categories. We visualise cluster structures and compare performance metrics to illustrate the benefits and limitations of each method. Finally, we analyse cluster composition relative to gender and socio-economic variables to highlight potential biases in the data and underscore the need for fairness-aware models and interpretability. By combining supervised and unsupervised techniques and explicitly addressing ethical considerations, this work contributes to more accurate, transparent and equitable ML-based depression diagnosis.},
 year = {2025}
}

Copy | Download

TY  - JOUR
T1  - Innovative Machine Learning Approaches for Accurate and Ethical Depression Diagnosis: Insights and Recommendations
AU  - Abiodun Abdulghani Akanbi
Y1  - 2025/12/19
PY  - 2025
N1  - https://doi.org/10.11648/j.mlr.20251002.17
DO  - 10.11648/j.mlr.20251002.17
T2  - Machine Learning Research
JF  - Machine Learning Research
JO  - Machine Learning Research
SP  - 158
EP  - 164
PB  - Science Publishing Group
SN  - 2637-5680
UR  - https://doi.org/10.11648/j.mlr.20251002.17
AB  - This study advances mental-health diagnostics by integrating supervised and unsupervised machine-learning methods with a strong ethical lens. We review and contextualize the current literature on ML applications for depression detection, noting the heavy reliance of supervised models on labelled data and the comparative under-exploration of unsupervised approaches. Using the public Depression Dataset-which comprises actigraph recordings from depressed patients and healthy controls and includes demographic and clinical attributes such as timestamps, activity counts, gender, age and Montgomery-Åsberg Depression Rating Scale  HYPERLINK "http://kaggle.com"  \h score we preprocess and engineer features capturing circadian rhythms and variability in motor activity. We then apply multiple clustering algorithms (K-means, hierarchical clustering and DBSCAN) to identify latent subgroups of depression severity, evaluating cluster validity via the silhouette score, Davies-Bouldin index and Calinski-Harabasz index. A supervised support-vector machine classifier trained on labelled severity categories serves as a baseline, and we find that unsupervised clustering achieves competitive performance while revealing nuanced patterns not captured by labelled categories. We visualise cluster structures and compare performance metrics to illustrate the benefits and limitations of each method. Finally, we analyse cluster composition relative to gender and socio-economic variables to highlight potential biases in the data and underscore the need for fairness-aware models and interpretability. By combining supervised and unsupervised techniques and explicitly addressing ethical considerations, this work contributes to more accurate, transparent and equitable ML-based depression diagnosis.
VL  - 10
IS  - 2
ER  -

Copy | Download

Author Information

Abiodun Abdulghani Akanbi

Department of Computer Science, Nile University of Nigeria, Abuja, Nigeria

Contact Email

http://orcid.org/0009-0006-7316-7014

Download PDF

Submit an Article

Sections

Plain Text BibTeX RIS

APA Style

Akanbi, A. A. (2025). Innovative Machine Learning Approaches for Accurate and Ethical Depression Diagnosis: Insights and Recommendations. Machine Learning Research, 10(2), 158-164. https://doi.org/10.11648/j.mlr.20251002.17

Copy | Download

ACS Style

Akanbi, A. A. Innovative Machine Learning Approaches for Accurate and Ethical Depression Diagnosis: Insights and Recommendations. Mach. Learn. Res. 2025, 10(2), 158-164. doi: 10.11648/j.mlr.20251002.17

Copy | Download

AMA Style

Akanbi AA. Innovative Machine Learning Approaches for Accurate and Ethical Depression Diagnosis: Insights and Recommendations. Mach Learn Res. 2025;10(2):158-164. doi: 10.11648/j.mlr.20251002.17

Copy | Download

@article{10.11648/j.mlr.20251002.17,
  author = {Abiodun Abdulghani Akanbi},
  title = {Innovative Machine Learning Approaches for Accurate and Ethical Depression Diagnosis: Insights and Recommendations},
  journal = {Machine Learning Research},
  volume = {10},
  number = {2},
  pages = {158-164},
  doi = {10.11648/j.mlr.20251002.17},
  url = {https://doi.org/10.11648/j.mlr.20251002.17},
  eprint = {https://article.sciencepublishinggroup.com/pdf/10.11648.j.mlr.20251002.17},
  abstract = {This study advances mental-health diagnostics by integrating supervised and unsupervised machine-learning methods with a strong ethical lens. We review and contextualize the current literature on ML applications for depression detection, noting the heavy reliance of supervised models on labelled data and the comparative under-exploration of unsupervised approaches. Using the public Depression Dataset-which comprises actigraph recordings from depressed patients and healthy controls and includes demographic and clinical attributes such as timestamps, activity counts, gender, age and Montgomery-Åsberg Depression Rating Scale  HYPERLINK "http://kaggle.com"  \h score we preprocess and engineer features capturing circadian rhythms and variability in motor activity. We then apply multiple clustering algorithms (K-means, hierarchical clustering and DBSCAN) to identify latent subgroups of depression severity, evaluating cluster validity via the silhouette score, Davies-Bouldin index and Calinski-Harabasz index. A supervised support-vector machine classifier trained on labelled severity categories serves as a baseline, and we find that unsupervised clustering achieves competitive performance while revealing nuanced patterns not captured by labelled categories. We visualise cluster structures and compare performance metrics to illustrate the benefits and limitations of each method. Finally, we analyse cluster composition relative to gender and socio-economic variables to highlight potential biases in the data and underscore the need for fairness-aware models and interpretability. By combining supervised and unsupervised techniques and explicitly addressing ethical considerations, this work contributes to more accurate, transparent and equitable ML-based depression diagnosis.},
 year = {2025}
}

Copy | Download

TY  - JOUR
T1  - Innovative Machine Learning Approaches for Accurate and Ethical Depression Diagnosis: Insights and Recommendations
AU  - Abiodun Abdulghani Akanbi
Y1  - 2025/12/19
PY  - 2025
N1  - https://doi.org/10.11648/j.mlr.20251002.17
DO  - 10.11648/j.mlr.20251002.17
T2  - Machine Learning Research
JF  - Machine Learning Research
JO  - Machine Learning Research
SP  - 158
EP  - 164
PB  - Science Publishing Group
SN  - 2637-5680
UR  - https://doi.org/10.11648/j.mlr.20251002.17
AB  - This study advances mental-health diagnostics by integrating supervised and unsupervised machine-learning methods with a strong ethical lens. We review and contextualize the current literature on ML applications for depression detection, noting the heavy reliance of supervised models on labelled data and the comparative under-exploration of unsupervised approaches. Using the public Depression Dataset-which comprises actigraph recordings from depressed patients and healthy controls and includes demographic and clinical attributes such as timestamps, activity counts, gender, age and Montgomery-Åsberg Depression Rating Scale  HYPERLINK "http://kaggle.com"  \h score we preprocess and engineer features capturing circadian rhythms and variability in motor activity. We then apply multiple clustering algorithms (K-means, hierarchical clustering and DBSCAN) to identify latent subgroups of depression severity, evaluating cluster validity via the silhouette score, Davies-Bouldin index and Calinski-Harabasz index. A supervised support-vector machine classifier trained on labelled severity categories serves as a baseline, and we find that unsupervised clustering achieves competitive performance while revealing nuanced patterns not captured by labelled categories. We visualise cluster structures and compare performance metrics to illustrate the benefits and limitations of each method. Finally, we analyse cluster composition relative to gender and socio-economic variables to highlight potential biases in the data and underscore the need for fairness-aware models and interpretability. By combining supervised and unsupervised techniques and explicitly addressing ethical considerations, this work contributes to more accurate, transparent and equitable ML-based depression diagnosis.
VL  - 10
IS  - 2
ER  -

Copy | Download