Glossary

Reinforcement Learning

What is Reinforcement Learning?

Reinforcement Learning uses the assignment of scores to action steps to maximize the total reward captured from a sequence of decisions. It is often used in game theory or decision trees to evaluate which steps will maximize the value of the final output. Like unsupervised machine learning, it does not require labeled data to evaluate the input variables. Different approaches within reinforced learning make different tradeoffs between exploration (evaluating different paths) and exploitation (using known scores) to maximize the end score.

Why is Reinforcement Learning important?

Reinforcement Learning applies to a number of use cases, especially when there is a model of the environment or scores can be discovered through simulating or interacting with the environment. In checkers or chess, for example, scores can be determined for a set number of moves ahead and evaluated to choose the next best move. Choosing investments can use Monte Carlo simulation to select the right diversified portfolio to maximize return for a given level of risk based on historical performance and variability.

How C3 AI Enables Organizations to Use Reinforcement Learning

C3 AI makes it easy to apply reinforcement learning to address domain-specific applications of AI to deliver business value today. The C3 AI Platform is a complete, end-to-end platform for designing, developing, deploying, and operating enterprise AI applications at industrial scale. The C3 AI ML Pipelines feature enables data scientists to easily build pipelines of different models to create complex analytic workflows, combining techniques from different domains to produce a more sophisticated result. For example, one could use image recognition to translate a complex diagram into its constituent components and labels, then match the labels to a database bill of materials, connect to available sensor data, build a model to predict potential failures hours or days in advance, and create work orders for provisioning replacement parts to address the issue during a maintenance window.

C3 AI also delivers a portfolio of prebuilt, SaaS enterprise AI applications for a growing number of use cases such as C3 AI Reliability, C3 AI Inventory Optimization, C3 AI Fraud Detection, C3 AI Anti-Money Laundering, and more. Some of the world’s largest organizations – such as Shell, US Department of Defense, Koch Industries, and others – use C3 AI technology to drive digital transformation initiatives that significantly reduce costs, increase asset availability and reliability, improve human safety, and enhance customer satisfaction.