Fig. 1From: Depression-level assessment from multi-lingual conversational speech data using acoustic and text featuresSample distribution before and after applying the ml-MRMR algorithm with Nmin=3. Turkish database is usedBack to article page