This tutorial covers the concept of training latent variable energy based models (LV-EBMs) and is is a part of the Deep Learning Course at CDS, a course that covered the latest techniques in deep learning and representation learning, focusing on supervised and unsupervised deep learning, embedding methods, metric learning, convolutional and recurrent nets, with applications to computer vision, natural language understanding, and speech recognition. Prerequisites for this course include: Introduction to Deep Learning, Parameter sharing, and Introduction to Data Science or a Graduate Level Machine Learning.
DAQCORD is a framework for the design, documentation and reporting of data curation methods in order to advance the scientific rigour, reproducibility and analysis of the data. This lecture covers the rationale for developing the framework, the process in which the framework was developed, and ends with a presentation of the framework. While the driving use case for DAQCORD was clinical traumatic brain injury research, the framework is applicable to clinical studies in other domains of clinical neuroscience research.
The Medical Informatics Platform (MIP) is a platform providing federated analytics for diagnosis and research in clinical neuroscience research. The federated analytics is possible thanks to a distributed engine that executes computations and transfers information between the members of the federation (hospital nodes). In this talk the speaker will describe the process of designing and implementing new analytical tools, i.e. statistical and machine learning algorithms. Mr. Sakellariou will further describe the environment in which these federated algorithms run, the challenges and the available tools, the principles that guide its design and the followed general methodology for each new algorithm. One of the most important challenges which are faced is to design these tools in a way that does not compromise the privacy of the clinical data involved. The speaker will show how to address the main questions when designing such algorithms: how to decompose and distribute the computations and what kind of information to exchange between nodes, in order to comply with the privacy constraint mentioned above. Finally, also the subject of validating these federated algorithms will be briefly touched.