Research on advanced algorithms for machine learning from data, with emphasis on advanced algorithms for machine learning from data, including traditional machine learning methods, modern deep neural networks, multi-task learning approaches, model interpretation techniques such as sensitivity analysis, SHAP analysis and other integrated approaches. The overall objective of the research is to contribute to both theory and applications with an effective design and use of advanced algorithms, methods and tools for machine learning, deep learning, deep neuroevolution, non-linear optimisation, sensitivity analysis, parameter optimisation. The application of algorithms and models may include different domains and data sources, classical problems and newer ones for deep learning, various real-world problems such as early diagnosis of neurodegenerative conditions from brain images (e.g. Alzheimer's disease), and/ or synthetic motor control problems for multiple tasks at humanoid scale.