Study: Algorithm can predict when an adolescent will become suicidal with 91% accuracy

Researchers from Brigham Young University, Johns Hopkins and Harvard person created an algorithm that tin foretell suicidal thoughts and behaviour among adolescents with 91% accuracy.

The researchers outline their instrumentality learning attack successful an nonfiction published contiguous successful PLOS ONE, wherever they besides item risk factors that are starring predictors of suicidal ideation and behavior among adolescents: online harassment and bullying.

"Suicide is the 2nd starring origin of decease among adolescents successful the U.S.," said Michael Barnes, survey coauthor and Associate Dean of the BYU College of Life Sciences. "It's captious we person a amended knowing of the hazard factors—and the protective factors—associated with this heartbreaking issue."

The survey results amusement researchers tin foretell with precocious accuracy which adolescents volition grounds suicidal thoughts (consider oregon planning) oregon suicidal behaviour (attempting) based connected experiences they face.

The squad analyzed information from 179,384 inferior precocious and high schoolhouse students, on with those who participated successful the Student Health and Risk Prevention survey from 2011-2017. The dataset includes responses to 300+ survey questions and 8000+ bits of demographic information, resulting successful a full of 1.2 cardinal data points that were processed. Researchers past applied assorted algorithms to the information and recovered a machine-learning exemplary that accurately predicted which adolescents went connected to person suicidal thoughts and behaviors (STB) based connected the information provided.

The information showed females were much apt to acquisition suicidal thoughts and behaviour (17.7%) than males (10.8%), and that those adolescents without a begetter successful the location were 72.6% much apt to person suicidal ideation than those that did.

Most importantly, the algorithm discovered which hazard factors were the starring predictors of suicidal thoughts and behavior:

• Being threatened oregon harassed done integer media
• Being picked connected oregon bullied by a pupil astatine school
• Exposure/involvement successful superior arguments and yelling astatine home

"This investigation finds the astir important basal causes of suicidal thoughts and behaviour successful adolescents and creates hazard profiles that springiness america a clearer representation of adolescents that are astatine risk," said survey coauthor Carl Hanson, prof of nationalist wellness astatine BYU. "If you privation to wrapper your caput astir what you tin bash astir it, these profiles are 1 bully spot to start."

Researchers were not amazed to spot immoderate of the hazard factors that roseate to the top—bullying and harassment—but were a spot taken to spot the dense power from family factors: 3 of the apical 10 predictive factors for STB were tied straight to household situations: 1) being successful a household wherever determination are superior arguments, 2) being successful a household that argues astir the aforesaid things implicit and implicit and 3) being successful a household that yells and insults each other.

The squad said the implications of the probe are captious for prevention programming and argumentation making. Specifically, they anticipation policymakers usage the STB risk illustration and its subordinate rankings to hole services, resources, and assessments aimed astatine school, assemblage and household settings.

"Clearly the results talk to the request for prevention and schools whitethorn beryllium the champion spot to commencement by helping to mitigate bullying and online harassment. The results besides bespeak a request to fortify families," Hanson said. "For communities, we request programming that tin assistance enactment and fortify the family."

---

---