Expert System based on Feature Selection for Classification of Self-Care Problems in Children and Youth People

One of the important problems for occupational therapists is self-care problems classification of children. The International Classification of Functioning, Disability, and Health for Children and Youth (ICF-CY) is a standard language to store the characteristics of child development in terms of health, education and social activities. Two expert models were proposed using Support Vector Machine classifier and also rule based classifier to predict the self-care problems of children and youth with physical and moor disability in earlier stage. The well-known and only available SCADI (Self-Care Activities Dataset based on ICF-CY) dataset was used. This dataset consists of numerous features. Generally, too many features take more space, time and not produce good result. Dimensionality reduction was used to select relevant features which are really needed for the models. The Boruta algorithm was used for feature selection. The Support Vector Machine model was developed using whole dataset and also 48 selected features in R Programming. The model outperformed with 97.14% on selected attributes rather than all attributes. This expert system will be useful for therapist to identify the problem of exceptional children.