Efficient or Powerful? Trade-offs Between Machine Learning and Deep Learning for Mental Illness Detection on Social Media

2025-03-03

AI Summary

This work addresses the trade-off among accuracy, interpretability, and computational efficiency in social media–based mental disorder detection. We systematically compare machine learning (ML) models—including logistic regression, LightGBM, and random forests—with deep learning (DL) approaches—namely ALBERT and GRU—on depression and anxiety classification tasks. Empirical evaluation on a medium-scale Chinese dataset reveals comparable F1-scores between ML and DL (differences ≤ ±1.2%), yet complementary strengths: ML models offer superior interpretability (e.g., feature attribution in logistic regression), 5–20× faster training, and >60% lower memory footprint; DL models better capture complex linguistic patterns, with ALBERT demonstrating optimal noise robustness. To our knowledge, this is the first model selection framework tailored to mental health NLP tasks, guiding practitioners in choosing appropriate models based on data scale, interpretability requirements, and computational constraints.

Technology Category

• Natural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models
• Machine Learning: Evaluation and Analysis
• Humans and AI: Human-in-the-loop Machine Learning

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Social media platforms provide valuable insights into mental health trends by capturing user-generated discussions on conditions such as depression, anxiety, and suicidal ideation. Machine learning (ML) and deep learning (DL) models have been increasingly applied to classify mental health conditions from textual data, but selecting the most effective model involves trade-offs in accuracy, interpretability, and computational efficiency. This study evaluates multiple ML models, including logistic regression, random forest, and LightGBM, alongside deep learning architectures such as ALBERT and Gated Recurrent Units (GRUs), for both binary and multi-class classification of mental health conditions. Our findings indicate that ML and DL models achieve comparable classification performance on medium-sized datasets, with ML models offering greater interpretability through variable importance scores, while DL models are more robust to complex linguistic patterns. Additionally, ML models require explicit feature engineering, whereas DL models learn hierarchical representations directly from text. Logistic regression provides the advantage of capturing both positive and negative associations between features and mental health conditions, whereas tree-based models prioritize decision-making power through split-based feature selection. This study offers empirical insights into the advantages and limitations of different modeling approaches and provides recommendations for selecting appropriate methods based on dataset size, interpretability needs, and computational constraints.

Problem

• Trade-offs between ML and DL for mental illness detection.
• Evaluating ML and DL models for mental health classification.
• Comparing accuracy, interpretability, and efficiency in mental health detection.

Innovation

• Compares ML and DL for mental health classification
• Evaluates models like logistic regression and ALBERT
• Highlights trade-offs in accuracy and interpretability