Improving Depression Prediction Accuracy Using Fisher Score-Based Feature Selection and Dynamic Ensemble Selection Approach Based on Acoustic Features of Speech

Abstract

Depression affects over 322 million people, and it is the most common source of disability worldwide. Literature in speech processing revealed that speech could be used for detecting depression. Depressed individuals exhibit varied acoustic characteristics compared to non-depressed. A four-staged machine learning classification system is developed to investigate the acoustic parameters to detect depression. Stage one uses speech recordings from a publicly available and clinically validated dataset DAIC-WOZ. The baseline acoustic feature vector, eGeMAPS, is extracted from the dataset in stage two. Adaptive synthetic (ADASYN) is performed along with data preprocessing to overcome the class imbalance. In stage three, we conducted feature selection (FS) using three techniques; Boruta FS, recursive feature elimination using support vector machine (SVM-RFE), and the fisher score-based FS. Experimentation with various machine learning base classifiers like gaussian naïve bayes (GNB), support vector machine (SVM), k-nearest neighbors (KNN), logistic regression (LR), and random forest classifier (RF) is performed in stage four. The hyperparameters of the classifiers are tuned using the GridSearchCV technique throughout the 10-fold stratified cross-validation (CV). Then we employed multiple dynamic ensemble selection of classifier algorithms (DES) with k=3 and k=5 utilizing the pool of aforementioned four base classifiers to improve the accuracy. We present a comparative study using eGeMAPS features against the base classifiers and the experimented DES classifiers. Our results on the DAIC-WOZ benchmark dataset suggested that K-Nearest Oracles Union (KNORA-U) DES with k=3 has superior accuracy using a subset of 15 features selected by fisher score-based FS than the individual base classifiers.

Enhancing Depression Prediction Accuracy Using Filter and Wrapper-Based Visual Feature Extraction

“…whereas our method does not need such clinical equipment's for data collection. In the study presented in [39], they used acoustic features and experimented with only few classifiers but lacks to include experiments with multiple classifiers. Our work analyses the performance over several classifiers using cost effective way of data collection.…”

Comparison of machine learning and logistic regression as predictive models for adverse maternal and neonatal outcomes of preeclampsia: A retrospective study

“…After the variable selection scheme was used to remove spurious variables, LR ( 26 ) and six ML prediction models were developed from the split training datasets, and the validation of models was tested on the testing datasets. GridSearch with Cross-Validation was the applied parameter optimization technique ( 27 ). The discriminative power of the models was assessed by the AUC of the ROC curve, and the calibration quality was determined by the calibration curve.…”

Analysis of Machine Learning Classifiers for Speaker Identification: A Study on SVM, Random Forest, KNN, and Decision Tree

“…The adoption of machine learning classifiers marked a significant shift in the research landscape of speaker identification. Studies have explored a variety of classifiers, including Support Vector Machines (SVM), Random Forests (RF), K-Nearest Neighbors (KNN), and Decision Trees (DT) (Jaid & AbdulHassan, 2023;Kumar & Das, 2022). SVMs have been noted for their effectiveness in high-dimensional spaces, with research demonstrating their capability in providing higher accuracy in speaker verification tasks (Alwahedi, Aldhaheri, Ferrag, Battah, & Tihanyi, 2024).…”