This research is implemented through the Operational Program Human Resources Development, Education and Lifelong Learning and is co-financed by the European Union (European Social Fund) and Greek national funds.
aiai folder is related with the corresponding publication submitted in AIAI2019 with title:
Investigating the benefits of exploiting incremental learners under Active Learning scheme - link
This paper tackles with the concept of applying Active Learning concept exploiting incremental learners, based mainly on SGD approach. Several binary datasets have been assessed and the results under 3 separate labeled-ratio-scenarios are really promising, regarding both accuracy and time efficiency.
Conference site: link
iisa folder is related with the corresponding publication submitted in IISA2019 with title:
Combining Active Learning with Self-train algorithm for classification of multimodal problems - link
This paper is oriented towards the combination of Uncertainty Sampling Query Strategy with Self-training scheme under a common framework, so as to obtain improved learning behavior against supervised mode, exploiting useful information from available data without relying heavily on the human oracle, through which trustworthy decisions are drawn for the most ambiguous unlabeled instances.
Conference site: link
More information are going to be mentioned after the presentations of these works on our academic site: http://ml.math.upatras.gr/