Detecting Depression Using an Ensemble Logistic Regression Model Based on Multiple Speech Features
Algorithm 2
ELRDD-E.
Input: training speech recordings of subject and its label and testing speech recordings of subject .
Output: depressed patient or healthy control label of subject g.
Step 1: call the training process of ELRDD; the inputs are and .
For k = 1 to 29
Step 2: call the testing process of ELRDD; the input is rk of subject and the outputs are probability pk for depressed patients and probability qk for healthy controls.
End
Step 3: , if the value of p is larger than q, subject is classified as depressed; otherwise, is classified as a control.