A deep Learning Approach to Malware detection in android platform
Throughout the year mobile devices such as tablets, smartphones and computers are extremely widespread because of the development of modern technology. By using these devices, users all over the globe can easily accessed a huge range of applications from both commercial and private use. Malware de...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Year Project Report |
| Language: | English English |
| Published: |
Universiti Malaysia Sarawak (UNIMAS)
2018
|
| Subjects: | |
| Online Access: | http://ir.unimas.my/id/eprint/29064/1/A%20deep%20Learning%20Approach%20to%20Malware%20detection%20in%20android%20platform%20%2824%20pgs%29.pdf http://ir.unimas.my/id/eprint/29064/2/A%20deep%20Learning%20Approach..ft.pdf http://ir.unimas.my/id/eprint/29064/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Throughout the year mobile devices such as tablets, smartphones and computers are extremely widespread because of the development of modern technology. By using these devices, users all over the globe can easily accessed a huge range of applications from both commercial and private use. Malware detection is an important aspect of software protection. As a matter of fact, the development of malware had begun soaring as more and more unknown malware were discovered. Malware is a common term used to describe malicious software that can induced security threats to any devices and also to the Internet network. In this study, a malware detection that is based on Deep Learning approach that utilize the Long-Short Term Memory Networks (LSTM) model in utilized. The chosen approach will learn and train itself by using the features that are needed for malware detection using a large data sets for evaluating the trained algorithm. The performance of the model is evaluated by comparing it with the Back-Propagation (BP) model. Results that was achieved by conducting the necessary experiments proved that the LSTM model is capable to detect malware with the error loss of 0.6 and achieved an accuracy of 93.60% compared to BP with an accuracy of 82.85%. |
|---|