The workshop on Life-Long Learning in Practical Applications is organized during the World Congress On Computational Intelligence in Padua, Italy.

About

Despite the proliferation on advanced machine learning and AI methods applied to plethora of applications, those approaches still suffer from problems and challenges such as: difficulties to adaptability to ever-changing data and data streams, the need for continuous improvement of the (once-established) models, missing and incomplete and unreliable data and datasets, the need to include human operator feedback in the loop, etc.

In this session we’d like to focus on practical applications of those new proposals and approaches to counter abovementioned problems such as e.g. life-long learning, online learning, transfer learning, active learning and other similar concepts.

We’d like to invite papers and projects focusing on those new approaches to machine learning in applied to practical problems such as: cybersecurity, anomaly detection, fake news detection, medical images analysis among others.

Therefore in this session, we would like to bring together researchers from ML and AI communities from various application areas who are working on practical implementations and deployment of new paradigms in machine learning, and on life-long learning systems and methods in particular.

Topics of interest

• learning from imbalanced data
• life-long learning methods and practical applications multi-task learning
• transfer learning
• zero-shot and few-shot learning
• online learning
• learning from drifted data streams
• semi-supervised learning
• self-labeling and active learning
• rare events detection
• learning from incomplete and missing data
• meta-learning
• dynamic neural architecture search
• recurrent concepts
• life-long learning methods and practical applications for:
  • anomaly detection,
  • cybersecurity
  • dissinformation detection
  • medical images analysis
  • industrial data processing and Industry 4..0
  • remote sensing and satellite imagery
  • intelligent assistants
  • chatbots
  • robotics
  • bioclimatic variables prediction
  • etc.
• benchmarking:
  • new benchmark problems for facilitating fair experimental comparisons and understanding performance among life-long learning methods.
  • proposal of new building blocks for an open-source, community-backed library for life-long learning.

Important dates

Paper submission

The instruction and link of submission system are available here If you would like to submit your work to our session please choose IJCNN track and select the L3-IPA special session

Session Chairs

• Sebastian Basterrech, VSB - Technical University of Ostrava, Czechia
• Michał Choraś, Bydgoszcz University of Science and Technology, Poland
• Jan Platoš, VSB - Technical University of Ostrava, Czechia
• Michał Woźniak, Wroclaw University of Science and Technology, Poland

Program Committee

TBA

Organizing Committee

• Filip Guzy
• Jędrzej Kozal
• Michał Leś
• Paweł Zyblewski