An Automatic Diagnosis of Mental Illness Using Optimized Dynamically Stabilized Recurrent Neural Network is proposed in this paper. The aim of the proposed method is " to automatically detect mental illness disor...
详细信息
An Automatic Diagnosis of Mental Illness Using Optimized Dynamically Stabilized Recurrent Neural Network is proposed in this paper. The aim of the proposed method is " to automatically detect mental illness disorder from OSMI data records " . The proposed approach automatically extracts the features needed for the network ' s training from the OSMI records in contrast to comparable literature research that employs conventional machine learning algorithms. Using the dual domain feature extraction method the time domain features (minimum, maximum, mean and medium) and statistic features (standard deviation, mean, kurtosis, and skew ness) of OSMI dataset is extracted. This research uses these extracted features for automatic diagnosis of mental illness disorder disease, and it has better accuracy performance. This is trained with Dynamically Stabilized Recurrent Neural Network (DSRNN) to extract key features in the dataset and train the network. By using OSMI dataset from different age groups, this research demonstrates a high success rate in categorizing mental illness and healthy individuals with 98 % and 99.5 % accuracy. DSRNN clearly shows the relation of frequency components among mental illness patients and the healthy individual. These results make it simple to discriminate between people with mental illnesses and healthy people.
暂无评论