Suggested Reading: Advances in News Recommendation

Major news sites like Google News or Yahoo! News as well as social media sites like Facebook or Twitter provide their users with personalized recommendations. These recommendations are tailored to the users’ individual reading preferences and are based on advanced machine learning techniques. Researchers at AAU Klagenfurt, TU Dortmund, and the University of Antwerp have recently published a survey on intelligent techniques and open-sourced a software framework for benchmarking such algorithms in a realistic setting.

Master Thesis Detection of alpine activities using Smartphones

Student: Christoph Lagger

Supervisor: Peter Schartner

Unfortunately accidents in alpine environments happen on a  daily basis, often during mountain hikes in summer or ski tours in winter. Besides  standardized security beacons (e.g. avalanche beep) everybody carries a smartphone with multiple sensors (such as Accelerometers and Gyroscopes among others) with them.  In emergency situations, time is crucial and an accurate and robust recognition system in form of a mobile application could trigger the chain of survival automatically and support rescue missions. In this thesis machine learning is used to determine current movement patterns or activities based on sensor data such as walking up/down, skiing down, pause, or in the worst case an emergency situation. We recorded a large dataset of actual movement patterns (7 days, 19 hours, 21 minutes and 22 seconds) from all available smartphone sensors during actual alpine activities. Movement data was analyzed and a comprehensive training dataset was created for further usage. The goal was to determine the best combination of sensors, algorithms, features and window size parameters to accurately detect said movement patterns. A framework was implemented to perform a series of experiments using 10-fold cross validation, evaluate its outcome and visualize movement data as well as simulate results. Evaluation results as well as simulation results showed that the Random Forest algorithm using data from the Gyroscope and Magnetometer sensor in combination with a 4-second sliding window and an overlap of 20%, utilizing the Root Mean Square, Mean, Signal Vector Magnitude, Energy, Variance, and Standard Deviation as features, achieved a promising F-Measure of 0.975.

Figure 1: Key activities and corresponding result of a simulation run using the most promising combination of algorithm, sensors, features and sliding window parameters. 

Master Thesis scan.net – Interactive Learning Platform for IT Security

Student: Andreas Schorn

Supervisor: Peter Schartner

 

Cyber security training is about training IT security experts and end users in the field of information security. Traditional teaching and learning methods, such as lectures and literature research, however, have been proven inadequate in the field of cyber security. Implementing basic security concepts in real-world environments is difficult for many people as they usually lack knowledge about the specific procedures. With the help of interactive exercises, an attempt is made in a practical way to implement these basic concepts in a realistic environment, and therefore facilitate better understanding of information security.

In this thesis an overview of different variants of cyber security training and cyber security exercises is given. Structure as well as implementation of such exercises, consisting of a secure exercise environment and hacking instructions, is explained in detail. The thesis contains approaches on how cyber security trainings can be implemented in higher education organisations and describes the development and evaluation of a cyber security training platform (scan.net) for lectures at the Alpen-Adria-Universität Klagenfurt.

 

 

 

 

 

 

 

Andrea Tonelleo was honored with the Aerospace Best Paper Award

Stochastic Trajectory Generation Using Particle Swarm Optimization for Quadrotor Unmanned Aerial Vehicles (UAVs) has been selected as the best research article published in 2017 in the MDPI Aerospace journal. The paper is co-authored by Babak Salamat and Andrea Tonello. It provides a realistic stochastic trajectory generation method for unmanned aerial vehicles. It offers a tool for the emulation of trajectories in typical flight scenarios, for instance, flight level, takeoff-mission-landing, and collision avoidance with complex maneuvering. The trajectories for these scenarios are implemented with quintic B-splines, which grants smoothness in the second-order derivatives of the Euler angles and accelerations. In order to tune the parameters of the quintic B-spline in the search space, a multi-objective optimization method called particle swarm optimization (PSO) is used. The proposed technique satisfies the constraints imposed by the configuration of the UAV. Further constraints can be introduced such as: obstacle avoidance, speed limitation, and actuator torque limitations due to the practical feasibility of the trajectories.

In the domain of aerial robotics, there is a large body of literature on path planning  and flight control. However, to assess performance, for instance of navigation algorithms, the trajectories followed by the moving aerial vehicle must be generated with a statistically representative emulator. In this paper, we have provided a new seminal idea on how to do so, and we believe that the results can open the door to a novel methodology to develop stochastic trajectory generator – prof. Tonello says.

Publications: Babak Salamat and Andrea M. Tonello. Stochastic trajectory generation using particle swarm optimization for quadrotor unmanned aerial vehicles (UAVs).  Aerospace 2017, 4(2), 27. Aerospace best paper awards 2017 – Editorial. Aerospace 2018, 5(2), 61.