Predictive Algorithms on AAL Context

Author: Prof Nuno Pombo STSM Period: 2019-10-28 – 2019-11-16

ECI: Yes

Hosting institution: Technical University of Košice

From ITC: Yes

 

PURPOSE OF THE STSM

The main objective of my Short Term Scientific Mission at the Technical University of Košice is to improve R&D capabilities on the AAL field, namely on the design of predictive algorithms for elderly monitoring and active ageing. Proposed contribution&Techniques: One of the main challenges is to design computerized systems capable to provide accurate and timely information that may support the clinical decision-making. On the other hand, the emergence of predictive analytics; primarily due to the rapid growth of data volumes, challenges for data handling techniques such as features selection, data imputation, and data fusion, just to mention a few, which benchmark and evaluation are a very promising challenge to be addressed. Finally, I intend to publish in co-authorship, at least one article on a scientific journal on the field of computer science. Discussing and learning I would like to build scientific relationships for a possibility of future collaborations.

Planning: 1.State of the Art: predictive algorithms [1st week, 2nd week] 2.Soft Skills: Teamwork strengthening [1st week, 2nd week] 3.Proof of concept: Application and evaluation of different methods [2nd week, 3td week] 4.Future work: planning future collaboration within SHELD-ON Cost Action, and draft version of joint publication [3td week].

DESCRIPTION OF WORK CARRIED OUT DURING THE STSMS

During the STSM, I got overview of the past and currently running projects and research activities of specific interest for the objectives of the CA16226 within the research groups of Prof. Frantisek Babic. The main purpose was to design a computerized systems capable to provide accurate and timely information that may support the clinical decision-making. Thus, we started to revise the state of the art, and to determine the functional, and non-functional requirements (such as performance criteria, data types, processing models, and so on) of the aimed system. In parallel, we presented and discuss our ongoing research. Later, we discuss on data sets cadidates to be used to verify and to validate the computerized system. The early stage experiments revealed the suitable of the system, however complementary studies are needed. Thus, the future cooperation will include additional experiments by a undergraduate student supervided by me, with the remote interaction of Prof. Frantisek Babic. The outcome of this activity will lead to a joint publication manuscript.

DESCRIPTION OF THE MAIN RESULTS OBTAINED

Within the STSM, several results were obtained from the design of the computerised model:

  1. System requirements;
  2. Data sets for system Veficication and Validation
  3. We paved the way for international cooperation on the topics covered by SHELD-ON COST Action

FUTURE COLLABORATIONS

This initial research will result in a joint journal publication in an ISI indexed journal and/or flagship conference. The future collaboration will include experiments with data sets provided by both institutions. We will try to  design a comprehensive set of algorithms over those data sets in order to enhance the current proposed system. In addition, Prof. Frantisek Babic will provide a webinar for students, and researchers of Assisted Living Computing and Telecommunications Lab (ALLab) at UBI.

Leave a comment