Por favor, use este identificador para citar o enlazar este ítem:
http://inaoe.repositorioinstitucional.mx/jspui/handle/1009/2157| Metaheuristic optimization approach for parameters estimation in arrhythmia classification from unbalanced data | |
| Juan Carlos Carrillo Alarcon | |
| Ignacio Algredo Badillo Luis A. Morales Rosales | |
| Acceso Abierto | |
| Atribución-NoComercial-SinDerivadas | |
| Electrocardiogram (ECG) Signal Processing Machine Learning Arrhythmia Unbalanced | |
| Cardiovascular diseases are a group of disorders or abnormalities in the functioning of the heart. These types of diseases are the leading causes of death in the world and Mexico. Hence, the detection of irregularities in the biosignals of the heart by electrocardiogram is vital. Arrhythmias are an example of cardiac diseases, causing irregularities in the cardiac electrical impulses. Due to the constant increase in this type of heart disease, specialists in cardiology are necessary. Unfortunately, not all hospitals or health clinics can have a cardiologist to treat these types of disorders. Therefore, it is important to design diagnostic tools that help local physicians to detect arrhythmias when accessing to a cardiologist is difficult. Several challenges exist in arrhythmia detection when computer-aided diagnosis systems are designed. The challenges are related to evaluating a large volume of data generated by heart monitors like Holter, the variability of ECG signal properties that fluctuate among patients, the noises or interference of external sources, and unbalanced arrhythmia datasets. A particular issue is the use of unbalanced datasets related to arrhythmias’ nature since these abnormalities do not regularly appear in the ECG signal. This problem affects the performance of computational models significantly for arrhythmias classification. Besides, in unbalanced-data contexts, where the classes are not equally represented, the optimization and configuration of the classification models are highly complex, reflecting on the use of computational resources. In this thesis, a metaheuristic-optimization approach is proposed by combining the data level and the algorithmic level to address the problem of data unbalance. We applied a four-stage methodology. The first stage focused on acquiring data and creating an unbalanced context of 8 types of arrhythmias. In the preprocessing stage, methods were applied to improve signal quality. In the third stage, features were extracted to obtain relevant information from the electrocardiogram signals. In the last phase, two metaheuristic approaches based on differential evolution and particle swarm optimization were implemented and compared. After testing in a specific context, the proposed approach was applied for multiple contexts, making variations to the dataset, such as class type, number of classes, and instances of each class. | |
| Instituto Nacional de Astrofísica, Óptica y Electrónica | |
| 2020-08 | |
| Tesis de maestría | |
| Inglés | |
| Estudiantes Investigadores Público en general | |
| Carrillo Alarcón, Juan Carlos., (2020), Metaheuristic optimization approach for parameters estimation in arrhythmia classification from unbalanced data, Tesis de Maestría, Instituto Nacional de Astrofísica, Óptica y Electrónica. | |
| INTELIGENCIA ARTIFICIAL | |
| Versión aceptada | |
| acceptedVersion - Versión aceptada | |
| Aparece en las colecciones: | Maestría en Ciencias Computacionales |
Cargar archivos:
| Fichero | Tamaño | Formato | |
|---|---|---|---|
| Tesis Completa Ing. Juan Carlos Carrillo Alarcón.pdf | 10.33 MB | Adobe PDF | Visualizar/Abrir |