Institutional Repository of Vilnius University
Mašininio mokymosi algoritmų bei dimensijos mažinimo taikymas plaučių ligų diagnostikoje /
Abstract
dc:descriptionThis paper examines dimensionality reduction influence on the classification accuracy. AlexNet and VGG16 convolutional neural networks were used to extract features from chest x-ray images and principal component analysis were applied to those extracted datasets. To classify data k-nearest neighbors algorithm, support vector machine and multilayer perceptron were used. A research was done by changing the amount of data to train classifiers. During the research, classification accuracy and training time results were collected. The research has shown that principal component analysis improves classification accuracy of classifiers trained with small amount of data and more than a few times reduces training time. In addition, the proposed method to concatenate features from AlexNet and VGG16 models has shown to improve classification accuracy compared to individual models.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2022
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Krikštaponis, Titas,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD146241430&prefLang=en_US