Diabetes Prediction using Machine Learning

• Published in: 2023 International Conference on Computer Communication and Informatics (ICCCI) pp. 1 - 10
• Main Authors: Parimala, G, Kayalvizhi, R., Nithiya, S.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 23.01.2023
• Subjects: and data interpretation; Biochemistry; Data models; Data visualization; Ensemble learning algorithms; machine learning; Machine learning algorithms; Medical services; Obesity; Prediction algorithms
• ISSN: 2473-7577
• DOI: 10.1109/ICCCI56745.2023.10128216

Diabetes is considered to be one of the worst illnesses in the world. Diabetes is caused by a combination of variables, including obesity, excessive blood glucose levels, and other causes. It does this by altering the insulin hormone, which in turn causes an irregular metabolism in the crab and raises its blood sugar levels. This program's primary objective is to lessen the risk that people may acquire diabetes by making forecasts for them and urging them to take more care of their diet and lifestyle in the years to come. The key goals of this research were to develop and execute a method for predicting diabetes using machine learning techniques, as well as investigate the strategies that would be used to achieve success in this Endeavour. The suggested technique makes use of a wide variety of classification and ensemble learning algorithms, some examples of which include Knn, Label Encoder, and train test split. The results of the research may provide information that will help medical professionals make more accurate early predictions and judgments in order to better manage diabetes and save lives. The method first extracts information from a dataset, such as certain symptoms that may be utilized to gain further knowledge about diabetes, and then validates that information using other data. This paper objective was to build classification models for the diabetes data set, develop models that can determine whether or not a person is sick, and get the greatest possible validation scores in the models that were developed. Massive datasets may be found in the healthcare business. By investigating enormous datasets in this manner, we may uncover previously unknown information and trends, which will enable us to draw conclusions based on the data and make accurate forecasts. We categorize the dataset using random techniques since our major goal in doing this research is to determine the method that is the most accurate for predicting diabetes. This will be accomplished by integrating machine learning, data visualization, and data interpretation. The use of machine learning, which is becoming more important in the modern healthcare sector, will be the focus of this research. Massive datasets may be found in the healthcare business.