Abstract
It is of great clinical significance to establish an accurate intelligent model to diagnose the somatization disorder of community correctional personnel. In this study, a novel machine learning framework is proposed to predict the severity of somatization disorder in community correction personnel. The core of this framework is to adopt the improved bacterial foraging optimization (IBFO) to optimize two key parameters (penalty coefficient and the kernel width) of a kernel extreme learning machine (KELM) and build an IBFO-based KELM (IBFO-KELM) for the diagnosis of somatization disorder patients. The main innovation point of the IBFO-KELM model is the introduction of opposition-based learning strategies in traditional bacteria foraging optimization, which increases the diversity of bacterial species, keeps a uniform distribution of individuals of initial population, and improves the convergence rate of the BFO optimization process as well as the probability of escaping from the local optimal solution. In order to verify the effectiveness of the method proposed in this study, a 10-fold cross-validation method based on data from a symptom self-assessment scale (SCL-90) is used to make comparison among IBFO-KELM, BFO-KELM (model based on the original bacterial foraging optimization model), GA-KELM (model based on genetic algorithm), PSO-KELM (model based on particle swarm optimization algorithm) and Grid-KELM (model based on grid search method). The experimental results show that the proposed IBFO-KELM prediction model has better performance than other methods in terms of classification accuracy, Matthews correlation coefficient (MCC), sensitivity and specificity. It can distinguish very well between severe somatization disorder and mild somatization and assist the psychological doctor with clinical diagnosis.
Highlights
Since the middle of the 20th century, all countries in the world have been reforming the penal system to explore more humane, scientific and effective punishment methods to improve the psychology and behavior of criminals
In order to improve the accuracy rate of assessment of the severity of the somatization disorder, this paper proposes a kernel extreme learning machine based on opposition-based Bacterial Foraging Optimization (BFO) method
A 10-fold cross-validation method is used on data in a symptom self-assessment scale (SCL-90) to make detailed comparison between improved bacterial foraging optimization (IBFO)-kernel extreme learning machine (KELM), BFO-KELM, GA-KELM, PSO-KELM and Grid-KELM
Summary
Since the middle of the 20th century, all countries in the world have been reforming the penal system to explore more humane, scientific and effective punishment methods to improve the psychology and behavior of criminals. In order to understand the psychological and behavioral characteristics of community correction objects, a self-rating scale (SCL-90) [1,2] is generally adopted to evaluate their mental health level. KELM has the multi-dominance of the ELM algorithm, and combines the kernel function to map the linear inseparable pattern non-linearly to the high-dimensional feature space in order to achieve linear separability and further improve the accuracy. In the newly-developed KELM, the introduction of a positive coefficient into the learning system makes the ELM more stable. If it is non-singular, the coefficient C can be added to the diagonal of HHT when calculating the output weight β.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.