|
Designing a Hybrid Model Using HSIC Lasso Feature Selection and Adaboost Classifier to Classify Image Data in Biomedicine |
|---|---|
| รหัสดีโอไอ | |
| Creator | Mukesh Madanan, Anita Venugopal |
| Title | Designing a Hybrid Model Using HSIC Lasso Feature Selection and Adaboost Classifier to Classify Image Data in Biomedicine |
| Contributor | - |
| Publisher | TuEngr Group |
| Publication Year | 2564 |
| Journal Title | International Transaction Journal of Engineering, Management, & Applied Sciences & Technologies |
| Journal Vol. | 12 |
| Journal No. | 1 |
| Page no. | 12A1G: 1-15 |
| Keyword | Artificial Intelligence, Medical Imaging, Object-Based Classification, Support Vector Machine (SVM), WEKA, AdaBoost SVM Classifier. |
| URL Website | http://TuEngr.com/Vol12_1.html |
| Website title | ITJEMAST V12(1) 2021 @ TuEngr.com |
| ISSN | 2228-9860 |
| Abstract | In cell-based research, an effective classification approach is required for visually monitoring a large quantity of image data of cells in vitro treatment. It is important to classify alive and dead cells likewise in tumor cell images, detecting virus-cell images, etc. to analyze patients' situation and then provide patient-centered care. Traditionally, the classification methods employed for classifying the cell microscopy data is time-consuming and is susceptible to faults and delusion. This is a serious and crucial dilemma. Accurate classification of data set is a major task in cell-based research as it determines the treatment. This paper introduces a hybrid model that uses a nonlinear HSIC Lasso feature selection method combined with the AdaBoost SVM Classifier to classify a large quantity of data effectively and efficiently. In the proposed model, object-based classification is executed within the bounds of the Waikato Environment for Knowledge Analysis (WEKA) interface. Besides, the accuracy of the classifier is evaluated by methods like feature selection and interactive learning in WEKA. The performance comparison of the proposed model amid existing classification approaches proved that the method is better in minimizing the mean absolute error value successfully. |
