Abstract:
The purpose of this paper is to develop a clear understanding of how to apply machine learning for the analysis of clinical data that can advance CIH, besides formal medicines. CIH methods are evolving into more common usage because of natural conservation practices, non-impactful techniques, and combined systematic approach. However, some of such treatments are helpful and safe hence requiring proper supervision and consultation. Some of the difficulties that were experienced during this research include lack of database which incorporates both conventional medicine and CIH practices. To cater for this, a special dataset was developed with the help of doctors and other medical experts. These files contain data about patients, symptoms they have and treatment that was provided to them during their stay in the health facility. Further, it covers data regarding the symptoms that are specific and the diseases that correspond with these particular symptoms. The first process of dealing with this information was data cleaning and data preprocessing and patient reports were collected, sorted and had Named Entity Recognition (NER) done on them with regard to symptoms. Finally, in order to extract the relevant treatment data the semantic analysis was carried out later. These steps are important so as to make sure that the clinical data is well formatted FOR analysis. Classification of models such as Support Vector Classification (SVC), Random Forest, Gradient Boosting, K-Nearest Neighbors (KNN),Neural Network, Multinomial Na¨ıve Bayes was carried out to determine the best CIH strategy in line with symptoms and history of the patients. There were high level of model validation and performance benchmarking using distance measurement of different coarse and treatments selection for text data. Thirdly, record information of patients greatly enhanced the reliability of the outcome prediction to aid constructed management strategies based on the illness and disease severity of patients. When additional patient data were included in the models, this led to better CIH treatment outcomes and more personalised therapy plans, proving the ability of machine learning in improving CIH practice.
