ED-Filter: Dynamic Feature Filtering for Eating Disorder Classification

TL;DR

This paper introduces ED-Filter, a novel dynamic feature selection method combining branch and bound search with deep learning, to improve eating disorder classification accuracy from Twitter data.

Contribution

The paper presents ED-Filter, a new feature selection approach that adapts to Twitter data's dynamic nature, enhancing ED classification accuracy over traditional methods.

Findings

ED-Filter outperforms conventional feature selection methods.

The hybrid algorithm effectively adapts to evolving Twitter data.

Significant improvements in ED detection accuracy on social media.

Abstract

Eating disorders (ED) are critical psychiatric problems that have alarmed the mental health community. Mental health professionals are increasingly recognizing the utility of data derived from social media platforms such as Twitter. However, high dimensionality and extensive feature sets of Twitter data present remarkable challenges for ED classification. To overcome these hurdles, we introduce a novel method, an informed branch and bound search technique known as ED-Filter. This strategy significantly improves the drawbacks of conventional feature selection algorithms such as filters and wrappers. ED-Filter iteratively identifies an optimal set of promising features that maximize the eating disorder classification accuracy. In order to adapt to the dynamic nature of Twitter ED data, we enhance the ED-Filter with a hybrid greedy-based deep learning algorithm. This algorithm swiftly identifies sub-optimal features to accommodate the ever-evolving data landscape. Experimental results on Twitter eating disorder data affirm the effectiveness and efficiency of ED-Filter. The method demonstrates significant improvements in classification accuracy and proves its value in eating disorder detection on social media platforms.