AHAdam Horvath

Adam Horvath

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Adam Horvath has a total of 6 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Shared depressive symptomatology across mental disorders: implications for mental health assessments

    by Adam Horvath, Jemma Todd

    Mental disorders with complex presentations often include depressive symptoms, which in turn might lead to misdiagnosis, contributing to disease burden statistics. The goal of the present study was to investigate whether a common depressive-symptom overlap was present in unstructured online forum posts, specifically on Reddit. The k-nearest neighbours voting classifier was fitted on posts downloaded from Reddit from seven mental-health support groups (depression, anxiety, bipolar, borderline personality, narcissistic personality, antisocial personality, schizophrenia). Phrases that contributed the most of the confusion were identified. The misclassification was asymmetrical: posts from four support groups (anxiety, bipolar, borderline personality, and schizophrenia) were misclassified often as being from the depression support group, but depression support group posts were not equally likely to be misclassified as being from another support group. Unstructured, unprompted self-disclosure datasets show high propensity of depressive symptoms, and strong convergent validity with existing theoretical frameworks and clinical findings.

  • Open Access

    Differential Diagnosis of Borderline Personality Disorder: Machine Learning Models of Subjective Experiences

    by Adam Horvath, Mark Dras, Miriam K. Forbes

    The present study aimed to capture the subjective experiences that distinguish BPD from closely related mental disorders to aid differential diagnosis. Posts from Reddit were downloaded from seven mental health discussion groups. The topics discussed in each post were extracted using a combination of machine learning approaches. Logistic regression models were used to classify whether posts originated from the BPD support group versus other support groups. The average classification performance was well above chance, even relying on only 25 subjective experiences, and after excluding topics that were objective markers of diagnoses (e.g., names of the disorders or medications). Fear of abandonment emerged as the key differentiator of BPD, and 11 of the 25 topics related directly to the definition of BPD in the DSM-5. However, several of these were typical of other disorders as well, raising questions about their diagnostic utility.


  • Predicting Suicidal Behavior Without Asking About Suicidal Ideation: Machine Learning and the Role of Borderline Personality Disorder Criteria

    by Adam Horvath, Mark Dras, Catie C.W. Lai, Simon Boag

    Objective: Identifying predictors contributing to suicide risk could help prevent suicides via targeted interventions. However, using only known risk factors may not yield accurate enough results. Furthermore, risk models typically rely on suicidal ideation, even though people often withhold this information.Method: This study examined the contribution of various predictors to the accuracy of six machine learning models for identifying suicidal behavior in a prison population (n = 353), including borderline personality disorder (BPD) and antisocial personality disorder (APD) criteria, and compared how excluding data about suicidal ideation affects accuracy.Results: Results revealed that gradient tree boosting accurately identified individuals with suicidal behavior, even without relying on questions about suicidal ideation (AUC = 0.875, F1 = 0.846). Furthermore, the model maintained this accuracy with only 29 predictors. Meeting five or more diagnostic criteria of BPD was an important risk factor for suicidal behavior. APD criteria, in the presence of other predictors, did not substantially improve accuracy. Additionally, it may be possible to implement a decision tree model to assess individuals at risk of suicide, without focusing upon suicidal ideation.Conclusions: These findings highlight that modern classification algorithms do not necessarily require information about suicidal ideation for modeling suicide and self‐harm behavior.