The Deep Learning Lecture Series 2020: DeepMind and the UCL Centre collaboration for Artificial Intelligence.
Deep learning plays a vital role in this epoch of artificial intelligence. Knowingly or unknowingly everyday we are utilising the windfalls of deep learning methods.Hence, therefore,DeepMind and the UCL Centre for Artificial Intelligence worked together to expedite a vision of facilitating a solid curative information on various significant concepts of deep learning through this series of lectures.
Introduction to Machine Learning and AI
This is the introductory part of this series, that constitutes the essential awareness about machine learning, deep learning concepts with the explanation of combining reinforcement learning and deep learning to achieve in building intelligent systems like AlphaGo, Capture the Flag and AlphaStar.
Neural Networks Foundations
The second video is all about the neural networks design, development and implementation. The video starts with the necessary information about how neural networks models operate, learn and solve the real-life problem.
Convolutional Neural Networks for Image Recognition
Here case studies of convolution neural network architectures are discussed elaborately. The speaker also emphasised the challenges in training deep learning models, techniques for finding the right architecture, and the building blocks of image recognition.
Vision Beyond Imagenet- Advanced Models for Computer Vision
This part is everything you need to know about the role of computer vision in building intelligent systems. It covers video processing with reinforcement learning strategies, and about self-supervised learning methodologies in uni-modal and multi-modal settings.
Optimisation for Machine Learning
This video explains the essentials of gradient-based optimisation techniques and their benefits towards training a neural network model.
Sequences and Recurrent Network
Here comes the final part. It uncovers the basics of sequence data modelling architectures. Also explains the basis of machine learning methods that are adapted to these specific structures.
References:
