Abstract
Background/objectives: Nowadays, there are thousands of approved drugs that can be used for treating people who have medical problems. Therefore, drug warnings and precautions are denoted to recognize a discrete set of adverse effects and other implied protection uncertainties that are useful for patient control. Methods/analysis/findings: In this study, the intended framework is divided into two principal stages: data retrieval and data processing. Firstly, in the data collection stage, drug reports, drug interactions, malfunctions, number of deaths, and other factors had been obtained from various references, including RxNorm and Drug Bank using web service. Secondly, in the data processing phase, different data mining algorithms used to classify drugs into suitable drugs and non-suitable drugs. Application/improvements: According to the experimental results, we found that the decision tree has more accuracy (97.9%) than other state-of-art methods. Keywords: Drug Interactions, Drugs Classification, Naïve Bayes, Support Vector Machine, Decision Tree.
Highlights
There is a comprehensive collection of different ways that have been used in healthcare supervision to complete the coordination process, which includes: J-48, Support Vector Machine (SVM), K-nearest neighbor, neural networks, Bayesian methods, etc
The C4.5 method is an expansion of the ID3 algorithm and is utilized to initialize a Decision Trees that can be applied for classification
In this sub-section, we comprise some of the research works that have been performed in the field of DDIs extraction and classification
Summary
Classification breaks data units into distinct groups. The categorization procedure foretells the aim class for several data points. Patients can be classified as “great danger” or “normal” patients with their illness model using data organization strategy. It is a supervised training procedure having identified class divisions. Only two available conditions such as, “true” or “false” danger inmate may be considered while the multiclass strategy has more than two purposes for example, “large,” “moderate,” and “fading” danger inmate [1,2,3]. The additional step is designing method where the assembled model that used for classification. There is a comprehensive collection of different ways that have been used in healthcare supervision to complete the coordination process, which includes: J-48, SVM, K-nearest neighbor, neural networks, Bayesian methods, etc
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
