Wits to host first Deep Learning Indaba in Africa
- Wits University
The Indaba will bring leaders in machine learning and artificial intelligence to Wits University to teach and mentor students, researchers and entrepreneurs.
The first Deep Learning Indaba will take place at the University of the Witwatersrand in Johannesburg, South Africa, from 10 – 15 September 2017.
Over five days, the Indaba will bring leaders in the fields of machine learning and artificial intelligence to Wits University to teach and mentor students, academics, researchers, technologists and entrepreneurs in the theory and practice of deep learning, one type of machine learning that uses deep neural networks and that is the basis of recent advances in text-to-speech systems, language translation and object recognition.
More than 300 attendees, hailing from 23 African countries, as well as from countries across the world, will participate in one of the largest machine learning teaching events globally. The Indaba aims to create a shared space to learn, teach, and to debate new developments in machine learning and artificial intelligence, and our African contributions to this scientific endeavour.
“Beyond the technical exchange, the Indaba will create opportunities for new research connections, to foster a better understanding of the variety of career paths in the field, and, we hope, through new friendships, perspectives and backgrounds, take the steps to realising a more diverse, racially-representative, and multicultural machine learning community,” says lead organiser of the Indaba, Shakir Mohamed, a Wits alumni and research scientist at DeepMind in the UK.
For Wits, the Indaba is an important opportunity to create a network of skilled individuals to solve problems and create new industries. “Remarkable centres of industry mushroom around strong, forward-looking research institutions such as Wits. There are many pressing challenges facing our continent and connecting with others throughout Africa and abroad helps the flow of knowledge into our country,” says Professor Ebrahim Momoniat, applied mathematician and the Dean of the Faculty of Science at Wits.
Speakers
The Indaba has attracted machine learning scientists from across the globe as speakers, including Nando de Freitas, Wits alumnus a lead research scientist at DeepMind. De Freitas is an authority in the field of machine learning, and in particular in the subfields of neural networks, Bayesian inference and Bayesian optimisation, and deep learning. He will be presenting a tutorial on aspects of convolutional neural networks at the Indaba.
George Konidaris, also a Wits alumnus, Assistant Professor at Brown University and Director of the Intelligent Robots Lab will also be teaching at the Indaba. Konidaris is a leading researcher at the intersection of robotics and machine learning and will be speaking about reinforcement learning, one of the topics in machine learning that addresses how artificial agents can learn from experience.
Richard Klein, Wits lecturer from the School of Computer Science and Applied Mathematics, will be speaking about recurrent neural networks. “My primary interest is in using machine learning and computer vision to support education and teaching endeavours. Using machine learning to visually understand how an audience is reacting to a speaker can help improve the quality of presentations and assist lecturers and presenters to better connect with their audience.”
Here is the full list of the local and international speakers and their topics.
Machine learning in Africa
Given the increasing focus on machine learning, the Indaba aims is to stimulate the participation of South Africans, and Africans more generally, within the research and innovation landscape surrounding deep learning and machine learning.
While African machine learning is a strong and varied field, African attendance and participation in internationally leading machine learning conferences is extremely low. This Indaba hopes to change as it is critical that Africans are the contributors, shapers and owners of the coming advances in machine learning.
Says Momoniat, “These are technologies that have the potential to empower our citizens and address many of the challenges facing our country and continent. The development of these skills needs to happen locally, to an international standard, so they can specifically target local problems. Conferences such as this are the key in forming and upskilling the local community of practitioners.”