Data fusion
| In general, we can say that "Data Fusion"
is any process of aggregating data from multiple sources into a single
composite with higher information quality. This is a recent topic
of research within CA3 and at this stage we mainly concentrate on
information fusion, which is a type of data fusion.
The goal of information fusion is to combine heterogeneous information
to obtain a single composite of potential comparable alternative
solutions that can be classified and ranked. The crux of information
fusion, which is a type of data fusion, is three-folded: (i) data
must be comparable and numerical, using some normalization process;
(ii) imprecision in data must be taken in consideration; (iii) an
appropriate aggregation function to combine values into a single
score must be selected.
Recently, the application of computational intelligence concepts
and techniques to perform data/information fusion is emerging as
a versatile tool and our research work follows this direction.
CA3 Relevant papers in this topic are (for more papers and downloads
see publication section):
- R. A. Ribeiro, A. Falcão, A.Mora, J. M. Fonseca. FIF:
A Fuzzy information fusion algorithm based on multi-criteria decision
making, Knowledge-Based Systems journal (2013).
- A. Mora, J. M. Fonseca, R. Ribeiro. Real-time Image Recovery
Using Temporal Image Fusion. IEEE Proceedings of the International
Conference on Fuzzy Systems, FUZZIEEE, India, 7-10 July (2013).
|
|
|
