Scholarlink

Purpose <p>The severity of depression among young Australians cannot be overstated, as it continues to have a profound impact on their mental health and general wellbeing. This study used machine learning (ML) algorithms to analyse longitudinal data, identifying key features to predict depression, assess future risk, and explore age-specific behaviours that contribute to its progression over time. The results emphasize the significance of early detection to prevent unfavourable consequences and shed light on the alterations in depressive symptoms during various stages of development.</p> Methods <p>Three widely regarded ML techniques—random forest (RF), support vector machine (SVM), and logistic regression (LR)—are being applied and compared with a longitudinal data analysis. Additionally, the Apriori algorithm is being utilized to explore potential relationships between health, behaviour, and activity issues with depression among different age groups (10–17).</p> Results <p>The analysis results indicate that the RF model is performing exceptionally well in diagnosing depression, with a 94% accuracy rate and weighted precision of 95% for non-depressed and 88% for depressed cases. In addition, the LR model shows promising results, achieving an 89% accuracy rate and 91% weighted precision. Moreover, insights from the Apriori algorithm underscore the significance of early detection by examining potential associations between health, behaviour, and activity problems and depression across diverse age groups.</p> Conclusion <p>Combining early screening programs with the RF model and the Apriori algorithm is crucial for understanding depression and developing effective prevention strategies. Emphasizing Apriori's factors and regularly updating strategies with new information will enhance depression management and prevention.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Insights into depression prediction, likelihood, and associations in children and adolescents: evidence from a 12-years study

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年2月28日

查看全文
  • Umme Marzia Haque,
  • Enamul Kabir,
  • Rasheda Khanam

摘要

Purpose

The severity of depression among young Australians cannot be overstated, as it continues to have a profound impact on their mental health and general wellbeing. This study used machine learning (ML) algorithms to analyse longitudinal data, identifying key features to predict depression, assess future risk, and explore age-specific behaviours that contribute to its progression over time. The results emphasize the significance of early detection to prevent unfavourable consequences and shed light on the alterations in depressive symptoms during various stages of development.

Methods

Three widely regarded ML techniques—random forest (RF), support vector machine (SVM), and logistic regression (LR)—are being applied and compared with a longitudinal data analysis. Additionally, the Apriori algorithm is being utilized to explore potential relationships between health, behaviour, and activity issues with depression among different age groups (10–17).

Results

The analysis results indicate that the RF model is performing exceptionally well in diagnosing depression, with a 94% accuracy rate and weighted precision of 95% for non-depressed and 88% for depressed cases. In addition, the LR model shows promising results, achieving an 89% accuracy rate and 91% weighted precision. Moreover, insights from the Apriori algorithm underscore the significance of early detection by examining potential associations between health, behaviour, and activity problems and depression across diverse age groups.

Conclusion

Combining early screening programs with the RF model and the Apriori algorithm is crucial for understanding depression and developing effective prevention strategies. Emphasizing Apriori's factors and regularly updating strategies with new information will enhance depression management and prevention.