Repository navigation
Iteration 3 Final project
batz91 edited this page Jun 8, 2018
·
2 revisions
In this iteration we want to present the final project.
- Finish last stages goals.
- Building a learning machine.
- Build a tool that simulates hacking into the software.
- Start leaning the machine about normal behavior of software
links to issues and boards
8 week
- Frida
- Visual studion- Unit test
- Aho corasick algorithm
- Pattern discovery algorithm
- Write the final report
This iteration was the most important iteration in the project. In this inertia, we construct a machine that learns with the help of different algorithms. We started tracking programs and identifying normal behavior patterns. By using the learned machine that we have built, we can identify at the time of running whether the behavior of a particular program is normal or abnormal. To better understand but look at our learning machine as a machine situations. To build it we have taken the following steps:
- We used the discovey patters algorithm to analyze program behavior after a long follow-up. This algorithm allows us to identify repetitive patterns and thus create a kind of language for a specific program. With the language we have created, we will be able to know in the future whether or not the program being tested is being used in words.
- At this stage after we created a language we used the Aho corasick algorithm to construct a state machine. We build the state machine with the words we found in the language. While program running, we analyze the activity of the program and we can know whether it uses words in this language by getting to the accept state in the state machine or alternatively using words that are not in the language, which indicates that we are not in accept state.