- Our paper on "An Approach to Stable Gradient Descent Adaptation of Higher-Order Neural Units" has been accepted to IEEE TNNLS.
- Courtesy of Prof. Ales Prochazka and DSP, ASPICC has taken a part in the organization of International Workshop on Computational Intelligence for Multimedia Understanding (IWCIM 2015) and in the closely following seminar on Real-time Image Guided Radiotherapy.
- ASPICC has started taking part in series of seminars and lectures: Digital Signal and Image Processing in Biomedical and Engineering Areas organized by DSP research group.
Seminars and Lectures
A Challenge: Development of Real-time Image Guided Radiotherapy - Tumor Tracking and Conformable Irradiation, 10:30-11:30
Prediction and Control of Respiration-Induced Lung Tumor Motion for Accurate Radiotherapy, 11:30-12:00
Nov 2, 2015, ASPICC, room A17, Fac. of Mechanical Eng., CTU in Prague
Lecture: Learning Entropy - an Approach to Novelty Detection , Oct 3, 2014, 13:45, A335 (VŠCHT, budova A)
Releases and Updates
AC-Kit renamed to ASPI Kit
AC-Kit 1.1 released (still a long way to go...)
The ASPICC focuses on research and implementations of Computational Intelligence with open-source IT technologies for:
- adaptive signal processing,
- real-data based modeling,
- adaptive control,
- real-time monitoring of dynamical systems,
- open-source and other modern IT technologies,
- prototyping applications,
- open-source consulting and innovation research.
The most common tasks that we aim are the typical ones for mechanical engineering field, but we also deal with multidiscplinary problems. Some research and applications topics of our focus can be listed as follows:
- neural networks including special classes of polynomial neural networks [1,2] (based on the cooperation with the University of Saskatchewan since 2003),
- novel algorithm for novelty detection in dynamical systems with the use of learning systems and multi-scale analysis – “Learning Entropy” [3, 4, 7] (based on the cooperation with the University of Manitoba since 2009),
- prediction and novelty detection in biomedical time series  also with applications of the Learning Entropy  (cooperation with Tohoku University since 2009),
- cooperation and networking with other departments of the Faculty of Mechanical Engineering, CTU in Prague, e.g.:
- monitoring and control of a railway stand,
- data processing and neural networks for energy consumption and heating,
- monitoring and optimization of low-power and high-power combustion processes,
- prediction of heat dilatation influences on accuracy of metal machinig, or
- multi-scale analysis of welding arc stability,
- also, we work on development of interactive web application and database applications with the use of open-source software tools (e.g. Django, OpenShift, Mezzanine)
- and some others (process data, Big and Gargantual Data)
Either within university projects or with industrial partners, the ASPICC group engages university students so they can learn and practice methods of Computational Intelligence, and consequently the students can contribute to the current research and evelopment.
Thus, graduate or even undergaduate students can learn some Computational Intelligence techniques and discover their potentials while they are also experiencing programming and implementation issues on real applications.
The ASPICC members are also active in cooperation with foreign universities, in engaging visiting researchers or exchange students, and in profesional and scientific comunities such as the IEEE Computational Intelligence Society).