• 12 min read
Conservative Q-Learning for Safe Sepsis Treatment
How offline reinforcement learning can save lives by learning safer treatment strategies without experimentation on patients.
Read essay
Thoughts on AI, healthcare, and building things that matter.