How to build an artificial intelligence system that can make its own decisions? Where can such systems be applied and why? Guests and participants of AI Conference will get answers to these and other questions from Vladimir Ivanov, Sr. Deep learning engineer at NVIDIA.
NVIDIA specializes in high-performance computing, development of processors and graphics accelerators. The American hardware developer is known in the market under such brands as Quadro, Tesla, ION and Tegra. The results of their work are used by scientists, artists and gamers, for whom 3D graphics and high power of the GPU are important.
The topic to be covered within the report of the NVIDIA representative Vladimir Ivanov is ‘Reinforcement learning (RL). Overview for beginners’. It is dedicated to the basics of machine learning with reinforcement with video examples.
- General formulation of the problem. To solve many problems, it is necessary to build a system that learns to make right decisions on its own. It can be used in robotics and games.
- How does RL differ from supervised and unsupervised learning? Video examples of RL application.
- Imitation learning. An example with autonomous driving, and how it is done in NVIDIA and Tesla. How to get and mark data.
- Simulations. If there is an effective simulation engine, then you can generate data and learn with its help.
- How to transfer an agent to the real world? Video examples.
The report of the deep learning expert Vladimir Ivanov will be held as part of the international event – AI Conference, which will take place on November 22 in Moscow.