The MAESTRA software is based on a tree-based and rule-based machine learning methods and is able to process extensive sets of data or streams of data, including even the incompletely labelled or network data, and give them a logical structure.
The software was already tested in a variety of fields. For example, their method was successfully applied to predict the phenotypes of micro-organisms from their genotypes and gene functions identified compounds to help treat tuberculosis and salmonella. Further, in the solar energy discipline, the MAESTRA methods were applied to predict both the production and the consumption of energy from different kinds of sensor data in different contexts. Similarly, Džeroski's team predicted equipment failures in trains and taxi demand from transport data, improved the accuracy of sentiment analysis and image annotation in social media, and worked on ML processes in the contexts of drug repurposing, tumor mutation, personalised medicine, brain informatics, sustainable food production and biodiversity.
For the future the MAESTRA team hopes for other users to further customise the tools (developed in open source) for commercial Artificial Intelligence's applications and add their own user interfaces. ‘This will allow MAESTRA partners to develop secondary products in the form of tools and services that are easier to use for potential customers,’ Džeroski says.
More information on this project is available on CORDIS :