Section: Computer Science and Engineering
Tutor: ALIPPI CESARE Major Research topic
:INTELLIGENT EMBEDDED SYSTEMS FOR HIGH-DIMENSIONAL AND HIGH-COMPLEXITY
Advisor: BORACCHI GIACOMOAbstract:
The main feature of intelligent systems is their ability to interact with the environment where they operate. Typically, these systems resort to data-driven models (e.g., classifiers) to address their tasks, and implicitly assume that the training data used to learn the model are from the same process that generates the test ones. Unfortunately, this stationarity is often an oversimplifying assumption because real-world processes typically change overtime. When a change occurs, the learned model becomes obsolete and the system performance might dramatically degrades. Therefore, an intelligent system is expected to adapt in non-stationary environments, by detecting changes in the process generating input data and updating its model accordingly.
STMicroelectronics is particularly interested in providing intelligent skills and adaptation mechanisms to their embedded systems, in particular in health and structure monitoring applications. One of the biggest challenges is that data acquired by these systems are often characterized by high dimensionality and high complexity, and they need to be handled by specific models that describe their structure. Such models must enable the prompt detection of process changes and must be easy to update considering the limited resources available on the embedded systems.
My research aims at developing a new intelligent data-processing methodology for analyzing and interpreting high-dimensional and high-complexity datastreams. The considered application scenarios are: health monitoring via wearable devices, structure monitoring by sensor networks, and quality control by visual inspection systems. \n\n