Practical insights into deep reinforcement learning

Sahika Genc dives deep into the current state-of-the-art techniques in deep reinforcement learning (DRL) for a variety of use cases.

Reinforcement learning (RL) is an advanced machine learning (ML) technique that makes short-term decisions while optimizing for a longer-term goal through trial and error. DRL is a subfield of RL that uses deep learning techniques to learn from raw sensor inputs, such as pixels from an image, without feature engineering to extract explicit information such as the borders of objects in an image. The cloud has lowered the barrier to entry through low-cost, highly scalable computing power, making DRL more accessible than ever to data scientists and developers in domains such as robotics, automation, and operations research.

Sahika explores the importance of DRL and how DRL agents learn. You’ll leave understanding how DRL agents accomplish challenges such as assisting manufacturing plant operators to program and repurpose industrial robotic platforms effectively and at a lower software cost; building managers to automatically adjust building temperature, ventilation, and lighting and reduce operating costs; and ecommerce providers to personalize product recommendations and adjust for rapidly changing trends.