人工智能
机器学习
深度学习
计算机科学
朴素贝叶斯分类器
分类器(UML)
推荐系统
逻辑回归
特征工程
支持向量机
出处
期刊:Journal of Artificial Intelligence and Copsule Networks
[Inventive Research Organization]
日期:2020-06-03
卷期号:2 (2): 121-130
被引量:47
标识
DOI:10.36548/jaicn.2020.2.005
摘要
There are several systems designed for the purpose of recommending. The recommending system has gained its prominence even in the medical industry for suggesting the diets for the patient’s, medicines to be taken, treatments to be taken etc. The recommendation system mainly enhances the robustness, extends protection against the many disease and improves the quality of living of an individual. So to automatically suggest the foods based on their health conditions and the level of sugar, blood pressure, protein, fat, cholesterol, age etc. the paper puts forth k-clique embedded deep learning classifier recommendation system for suggesting the diets for the patients. The K-clique incorporated in the recommendation system in an effort of getting an improved preciseness and increasing the accuracy of the deep learning classifier (gated recurrent units). The dataset for the empirical analysis of the developed system was performed with the data set of the patients collected over the internet as well as hospitals, information’s of about 50 patients were collected with thirteen features of various disease and thousand products with eight feature set. All these features were encoded and grouped into several clusters before applying into the deep learning classifiers. The better preciseness and the accuracy observed for the developed system experimentally is compared with the machine learning techniques such as logistic regression and Naïve Bayes and other deep learning classifiers such as the MLP and RNN to demonstrate the proficiency of the K-clique deep learning classifier based recommendation system (K-DLRS)
科研通智能强力驱动
Strongly Powered by AbleSci AI