Home » Corsi » Traffic Classification with Statistical Methods

 
 

Traffic Classification with Statistical Methods

 
 

Held by: Dr. Riccardo Pecori.

Internet traffic classification has moved in the last years from traditional port and payload-based approaches towards methods employing statistical measurements and machine learning techniques.
The course aims to provide, first, the basics of traffic classification, i.e., information about the most used tools, past and current classifiers, the most used datasets and identified protocols as well as the level of granularity employed. Secondly, some machine learning and data mining techniques applied to traffic classification are briefly reviewed, and some significant use cases from the literature are analyzed.

Finally, current research topics such as the utilization of side information and traffic classification applied to anonymity systems are discussed.
A final test must be passed to get 1 CFU.

Program:
- Lecture 1: Basics of Traffic Classification
Motivations, tools, types, datasets, protocols, features. Typologies, performances and limitations of some well-known statistical methods.
- Lecture 2: Literature relevant examples Literature examples and performances of applying machine learning techniques to different traffic classification scenarios.
- Lecture 3: Ongoing Research Topics Exploitation of side information, ground truth analysis and validation, applications to anonymity
systems. Final test.

Schedule:

June 20th, July 4th, July 11th 2017. Time 15:00-17:00

CFU:

0,5 CFU for attending, 1 CFU for attending and passing the final exam.

Comments are closed

Sorry, but you cannot leave a comment for this post.

News »

Welcome to our PhD program – XXXII cycle

Ci spiace, ma questo articolo รจ disponibile soltanto in English.

 
 
 
English Italy
italyEnglish