Applied Reinforcement : Learning Business optimization and LLM fine-tuning ( MEAP v3 )

Optimize business processes, people, and resources using AI and the power of reinforcement learning.

Whether you’re finding the best delivery route, establishing efficient schedules, or maximizing profit with dynamic pricing, success in business can come down to the right optimizations. AI tools like large language models can help. When you tune them to your specific business data, they really start to shine! Applied Reinforcement Learning teaches the essentials of business optimization using reinforcement learning and AI models through relevant and useful business applications. You’ll apply RL to supply chains, marketing and ad campaigns, logistics, and even optimizing AI chatbots. Graphics, code samples, and math-lite explanations demonstrate the core theories of RL in an intuitive and illustrative way.

In Applied Reinforcement Learning you’ll learn
RL for real-world challenges like scheduling, routing, and pricing
Custom simulation environments to train RL agents
RL algorithms including contextual bandits, Deep Q-Networks and actor-critic methods
End-to-end problems for e-commerce, vehicle routing, and supply chain management
Integrate RL with large language models using RLHF

Reinforcement learning models develop through trial and error, exploring their environment and learning from successes and mistakes. This powerful AI approach can easily be turned to automatically optimizing business processes like pricing, logistics, and customer engagement. In Applied Reinforcement Learning you’ll bring RL to solve common yet practical industry challenges. You’ll discover both the algorithms that underpin RL and how to build the simulation environments you’ll need to train custom models.

about the book
Applied Reinforcement Learning presents RL in an intuitive way, effectively applying this powerful technique in real-world environments. Each chapter explores an end-to-end industry case study—including optimizing an ad campaign using contextual bandit algorithms, production line scheduling problems using tabular RL and Deep Q-Networks for real-world business challenges, and applying dynamic pricing with Deep Deterministic Policy Gradient for solving dynamic pricing problems. For each example, you’ll step into the role of a consultant, analyzing how a problem can be effectively solved with RL. You’ll discover full coverage of the latest and most relevant techniques for RL, including utilizing reinforcement learning with human feedback (RLHF) to align large language models into business objectives and constraints.

about the reader
For readers comfortable with business processes and intermediate level programming. No advanced math or specialist AI knowledge is required.

about the author
Hadi Aghazadeh is a Machine Learning Engineer at Bits in Glass, where he applies advanced AI and generative AI solutions to real-world business challenges. He has delivered numerous high-impact projects—from dynamic pricing in ride-hailing to fraud detection in energy and banking. Hadi has earned multiple awards, including first place in the Alberta Machine Intelligence Institute Reinforcement Learning Competition and the prestigious Alberta Innovates Scholarship.