Survey of scientific projects in applied AI

The questionnaires about 63 projects collected by partner organizations from the 5 countries were collected and analyzed, concerning artificial intelligence teaching. The project coordinators were from 19 countries. Among the more interesting results belongs the finding, that more than half of the projects concerned deep neural networks learning modules, and most machine learning tasks which were solved, were image processing, classification, regression, clusterization and natural language processing. Among the used AI libraries dominated TensorFlow, Keras, scikit-learn and CUDA. The programming languages were Python and C++.

Most of the answers were analyzed and visualized in a form of graphs.

Full document can be found here: Research 3 StateOfTheArt_scientificprojects_Dirgova