An Intelligent Disease Prediction System for Psychological Diseases by implementing Hybrid Hopfield Recurrent Neural Network Approach

Vikas Kamra, Praveen Kumar, Masoud Mohammadian

Research output: Working paperPreprint


Machine learning approaches for automatic disease prediction from amorphous clinical data items has become a popular study topic in recent years. Existing electronic health record analysis work in the medical area improves early-stage illness diagnosis by substantiating relevant information connected to patients with disease in huge quantities. However, when standard rule-based approaches are employed, which are unable to handle amorphous clinical data items, the advantages of analysis are not accomplished adequately. Because the utilization of all text written in doctor’s prescription is not equal in the detection of diseases, a single technique will not be able to manage all problems associated with the analysis of amorphous data. As a result, there is a requirement to build a hybrid-model that solves these potential difficulties, which is an intriguing subject matter that should be investigated further. In light of the aforementioned issues, this study offers a hybrid Hopfield recurrent neural network (H2RN2) technique based on amorphous clinical data items taken from Kaggle database. Using fivefold cross validation technique within a recurrent neural network detracts over fitting of the model. The proposed model automatically learns inherent semantic characteristics from available clinical data items. In addition to effective learning during training of the model, the hybrid approach also helps in accurate prediction of the disease with improved accuracy. The developed model is assessed using three measuring parameters, accuracy, recall and F1 score while using available amorphous clinical dataset. The proposed model yielded an accuracy of 97.53% in experimental evaluation, which is superior to several existing approaches for psychological diseases. The hybrid Hopfield recurrent neural network (H2RN2) approach for prediction of psychological diseases is presented in this paper. The results demonstrate that the proposed model outperforms several current techniques in predicting the risk of psychiatric disorders. The similar approach may be used to anticipate the risk of different diseases.
Original languageEnglish
Place of PublicationUnited Kingdom
Number of pages21
Publication statusIn preparation - 31 Jul 2022

Publication series

NameIntelligent Systems with Applications
ISSN (Print)2667-3053

Cite this