Comparing machine learning to a rule-based approach for predicting suicidal behavior among adolescents: Results from a longitudinal population-based survey

Vuuren, C.L. van; Mens, Kasper van; Beurs, Derek de; Lokkerbol, Joran; Wal, Marcel F. van der; Cuijpers, Pim; Chinapaw, Mai

doi:10.1016/j.jad.2021.09.018

Cited by 18 publications

(10 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The fine-tuned model reached an excellent AUC of 92.1% ( 61 ). This figure is near the highest AUC reported in existing literature, ranging from 0.716 to 0.925 ( 23 , 25 , 26 , 28 ). With a high specificity of 0.90 and a high sensitivity of 0.77, the model could correctly classify 77% of adolescents reporting a recent suicide attempt as well as 90% of adolescents not reporting a recent suicide attempt.…”

Section: Discussionsupporting

confidence: 54%

“…However, modifying the underlying distribution of the outcome breaches the principal machine learning assumption that the training and testing datasets are sampled from the same population ( 45 ). To resolve this issue, previous machine learning studies on suicide attempt prediction have often balanced the testing dataset as well [see for example, ( 26 , 29 )], rendering the model inapplicable to the real-world problem. Simply put, when the distributions of both the training and testing datasets differ from the underlying population distribution, the performance of the model on a real-world dataset with severe class imbalance would be unknown ( 45 ).…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Predicting suicide attempts among Norwegian adolescents without using suicide-related items: a machine learning approach

Haghish,

Czajkowski,

von Soest

2023

Front. Psychiatry

View full text Add to dashboard Cite

IntroductionResearch on the classification models of suicide attempts has predominantly depended on the collection of sensitive data related to suicide. Gathering this type of information at the population level can be challenging, especially when it pertains to adolescents. We addressed two main objectives: (1) the feasibility of classifying adolescents at high risk of attempting suicide without relying on specific suicide-related survey items such as history of suicide attempts, suicide plan, or suicide ideation, and (2) identifying the most important predictors of suicide attempts among adolescents.MethodsNationwide survey data from 173,664 Norwegian adolescents (ages 13–18) were utilized to train a binary classification model, using 169 questionnaire items. The Extreme Gradient Boosting (XGBoost) algorithm was fine-tuned to classify adolescent suicide attempts, and the most important predictors were identified.ResultsXGBoost achieved a sensitivity of 77% with a specificity of 90%, and an AUC of 92.1% and an AUPRC of 47.1%. A coherent set of predictors in the domains of internalizing problems, substance use, interpersonal relationships, and victimization were pinpointed as the most important items related to recent suicide attempts.ConclusionThis study underscores the potential of machine learning for screening adolescent suicide attempts on a population scale without requiring sensitive suicide-related survey items. Future research investigating the etiology of suicidal behavior may direct particular attention to internalizing problems, interpersonal relationships, victimization, and substance use.

show abstract

Section: Discussionsupporting

confidence: 54%

Section: Methodsmentioning

confidence: 99%

Predicting suicide attempts among Norwegian adolescents without using suicide-related items: a machine learning approach

Haghish,

Czajkowski,

von Soest

2023

Front. Psychiatry

View full text Add to dashboard Cite

show abstract

“…Moreover, the process of variable selection enables models to achieve higher accuracy and better generalization capabilities. For example, van Vuuren and colleagues [20] found that LASSO created a model that was able to classify students as at risk for suicide with a higher accuracy than simple inclusion rules (i.e., predicting based on history of suicide alone). Pratik and colleagues [21] utilized Elastic Net to select variables that were able to predict smoking addiction in young adults with higher accuracy than previous research.…”

Section: Variable Selection In Machine Learningmentioning

confidence: 99%

A Tutorial on Supervised Machine Learning Variable Selection Methods for the Social and Health Sciences in R

Bain,

Shi,

Ethridge

et al. 2024

Preprint

View full text Add to dashboard Cite

With recent increases in the size of datasets currently available in the behavioral and health sciences, the need for efficient and effective variable selection techniques has increased. A plethora of techniques exist, yet only a few are used within the psychological sciences (e.g., stepwise regression, which is most common, the LASSO, and Elastic Net). The purpose of this tutorial is to increase awareness of the various variable selection methods available in the popular statistical software R, and guide researchers through how each method can be used to select variables in the context of classification using a recent survey-based assessment of misophonia. Specifically, readers will learn about how to implement and interpret results from the LASSO, Elastic Net, a penalized SVM classifier, an implementation of random forest, and the genetic algorithm. The associated code and data implemented in this tutorial are available on OSF to allow for a more interactive experience. This paper is written with the assumption that individuals have at least a basic understanding of R.

show abstract

“…In the following paragraphs, we will summarize three major points: First, the most important variables across the three models by far were indicators of previous self-harm. The question arises as to what extent algorithms that can incorporate several hundreds of variables have incremental value over a simple decision rule that classifies every adolescent who ever showed previous self-harming behavior as ''at risk'' (e.g., Van Vuuren et al, 2021). Using such a single-item decision rule (i.e., classifying every 17-year-old who confirmed previous self-harm at 14 years of age as ''at risk''), balanced accuracy was only slightly lower than that in the first model (.74 vs. .76), but sensitivity was substantially lower (.59 vs .69).…”

Section: Important Variables In Predicting Suicide Attemptsmentioning

confidence: 99%

Predicting Lifetime Suicide Attempts in a Community Sample of Adolescents Using Machine Learning Algorithms

2023

View full text Add to dashboard Cite

Suicide is a major global health concern and a prominent cause of death in adolescents. Previous research on suicide prediction has mainly focused on clinical or adult samples. To prevent suicides at an early stage, however, it is important to screen for risk factors in a community sample of adolescents. We compared the accuracy of logistic regressions, elastic net regressions, and gradient boosting machines in predicting suicide attempts by 17-year-olds in the Millennium Cohort Study ( N = 7,347), combining a large set of self- and other-reported variables from different categories. Both machine learning algorithms outperformed logistic regressions and achieved similar balanced accuracies (.76 when using data 3 years before the self-reported lifetime suicide attempts and .85 when using data from the same measurement wave). We identified essential variables that should be considered when screening for suicidal behavior. Finally, we discuss the usefulness of complex machine learning models in suicide prediction.

show abstract

Comparing machine learning to a rule-based approach for predicting suicidal behavior among adolescents: Results from a longitudinal population-based survey

Cited by 18 publications

References 26 publications

Predicting suicide attempts among Norwegian adolescents without using suicide-related items: a machine learning approach

Predicting suicide attempts among Norwegian adolescents without using suicide-related items: a machine learning approach

A Tutorial on Supervised Machine Learning Variable Selection Methods for the Social and Health Sciences in R

Predicting Lifetime Suicide Attempts in a Community Sample of Adolescents Using Machine Learning Algorithms

Contact Info

Product

Resources

About