Predictive analysis of stroke cases

Yu Feng Gao

Bachelor's thesis

Abstract:

The theoretical part of the thesis begins with our definition of stroke and the combined influencing factors that lead to it. The concept of big data is then explained and we introduce the basic concepts and principles of logistic regression algorithms for big data, before focusing on the development of statistical data processing and the basic methods of statistical and predictive analysis. In the practical part of the work, we dealt with a large dataset of 5100 objects, a collection containing patient-related information placed on the Kaggle online platform for educational purposes. The practical work involved creating appropriate classification algorithms and using analytical tools to identify the factors that influence stroke. Based on these factors, logistic regression models and decision trees were constructed to predict strokes. Finally, the predictive power of the model was validated using the analytical tools and the application of new data. Two data prediction models were developed using logistic regression and decision tree algorithms encapsulated in SAS studio to model data based on a dataset of stroke patients. The two data models show that four factors - age, average blood glucose level, hypertension and heart disease - are risk factors that influence stroke. Of these, age is the most important influencing factor. By comparing the two prediction models, a data model with a prediction accuracy of 84% was created, which can provide a reliable reference for stroke prediction.

Keywords

Big data, Stroke, Statistical Analysis, Predictive Modeling

Language used: English

Date on which the thesis was submitted / produced: 15. 3. 2022

Thesis defence

• Supervisor: Ing. Tomáš Hlavsa, Ph.D.

Czech University of Life Sciences Prague

Faculty of Economics and Management

Bachelor programme:
Economics and Management