Article Preview
Top1. Introduction
Doctors are trained based on traditional knowledge for treatment purposes with insufficient studies, and they acquire deep implicit knowledge after years of studies. However, the demands for diabetes mellitus are high and possibly outside the knowledge from books. Diabetes mellitus patients are faced with metabolic health problems, neuropathy, skin condition and many more. An individual diagnosed with diabetes patients is uncovered to hyperglycaemias for many years and out coming is at higher risk for diabetes difficulties. In addition, this traditional process takes a long time. Diabetes prediction using machine learning helps to identify the patterns much earlier. According to Alghamdi et al. (2017), machine learning is a difficult subject with popularity because it can serve as a general-purpose programming language and its adoption in both scientific computing and artificial intelligence. Often this study requires the support of use cases and their applications in different domains or disciplines.
Machine learning is used for predicting diabetes and getting preferable outcomes. Decision trees are a well-liked machine learning technique in the medical area and effectively provide probabilities of key decisions. In this regard, Neural network algorithms are currently a common machine learning approach that performs better. The support vector machine, the Decision Tree, Naive Bayes algorithm and K-nearest algorithm, and the Random Forest (RF) algorithm have been discussed in this paper. The random forest model assumes a large number of decision trees, and those are used to predict diabetes. Here, this report used a diabete.csv dataset that helps to predict the diabetes rate.
The research aim of the project is to design a high-accuracy diabetes prediction application through machine learning algorithms, which can help identify patterns earlier. The primary goals of this study are as follows:
- •
To construct a web application that will allow users to evaluate whether or not a patient has diabetes.
- •
To use the algorithms of SVM, Decision Tree and Random Forest to manipulate the data.
- •
To critically visualize the data set and find out the correlation between several features of the data set.
- •
To develop a general workflow of the predicted outcomes To evaluate the outcomes of the end products critically.
This paper is organized as follows: Section 2 reviews the literature on diabetes prediction and machine learning algorithms. Section 3 introduces the methods used in this research. Section 4 reports the results after conducting the algorithms, followed by the discussion on these results in section 5. Finally, we conclude our research in Section 6.