Ruijter, R.J.A. de (2017) The Effect of pooling the Hidden Activation in a Patch-Based Image Classification System using Multi-Layer Perceptrons. Bachelor's Thesis, Artificial Intelligence.
|
Text
AI_BA_2017_Rogier_de_Ruijter.pdf - Published Version Download (420kB) | Preview |
|
Text
Toestemming.pdf - Other Restricted to Backend only Download (77kB) |
Abstract
In this thesis a patch-based image classification system using a multi-layer perceptron (MLP) system is proposed. The patch-based MLP is trained on a set of randomly selected patches from the training images. This method is also used in a design where multiple patch-based MLPs are trained on specific regions of the images, this is called the scoped patch-based MLP system. Both the patch-based MLP and the scoped patch-based MLP system are used as a feature extractor for a classifier as well. The performance of these approaches are tested on the MNIST and CIFAR-10 datasets, where 99.43% and 71.63% classification accuracy are obtained, respectively.
Item Type: | Thesis (Bachelor's Thesis) |
---|---|
Degree programme: | Artificial Intelligence |
Thesis type: | Bachelor's Thesis |
Language: | English |
Date Deposited: | 15 Feb 2018 08:26 |
Last Modified: | 15 Feb 2018 08:26 |
URI: | https://fse.studenttheses.ub.rug.nl/id/eprint/14828 |
Actions (login required)
View Item |