But more than that, we need the ability to interpret the numerical data compiled by machines and convert it into formats that People can understand and use. However, traditional methods of learning, mining and visualization face unprecedented challenges and, by the same token, great opportunities, as the volume of data and computational capacity continues to grow.
This research area will focus on emerging topics related to learning, mining and visualization, seeking:
Address issues related to data robustness and quality, to improve the quality, structure and scope of data training, which are key points in the learning process.
Use learning to detect conflicts, anomalies or low credibility information.
Propose new visualizations and interactive media that transform the experience of machines into forms that People can explore, conceptualize, share and use.
Strengthen contemporary learning models in the context of deep learning, designing new strategies for sequential learning, reinforcement and imitation to take full advantage of available data. In addition, in relation to social networks, generate mechanisms to deal with data loss or bias.
