ICI Project Coordinator: Dr. Eng. Liviu BADEA – Scientific Researcher I

Overall objective of the project

The project aimed to develop advanced research on convergence of methods and techniques used in Big Data, intelligent data analysis ("Analytics") and advanced learning ("Deep Learning").

Project description

The project started from the idea of ​​complementarity, synergy and convergence between Big Data and Deep Learning, truly exploited with the latest developments – interrelated – in both areas. Until recently, the Big Data approach has focused particularly on data management. These are extremely useful in many applications, but in many circumstances there is still much information left out of date. In this case, automated learning or "data mining" techniques are useful. However, automated learning, due to the complexity of specific algorithms, has so far approached relatively smaller sets of relatively small data sets. Methods such as Deep Learning and Big Data Analytics have recently been developed, applicable on complex and very large data sets (millions of instances / examples).

The project explores advanced data management, analysis and learning technologies for Big Data, researching and developing algorithms and planning methods for data intensive platforms, and addressing complex issues with these technologies and developing the appropriate expertise that can be based on future developments, projects and applications.

Results

• study / analysis of the Big Data Analytics and Deep Learning domains;
• Apache Spark based platform, support for advanced geospatial data analysis (experimental model);
• task schedule for intensive platforms as data; simulation and evaluation environment for this model (experimental model);
• complex applications based on deep convolutional neural network architectures, object classification in images application, and histopathological imaging analysis (experimental model).