Answers are grounded in this paper and cite their sources.
Pricing techniques and strategies are changing and developing within the domain of e-commerce and retail. These sectors implement differentiated pricing systems depending on market conditions, industry competition, and product or service demand. Market participants that apply a dynamic pricing technique also face challenges associated with traditional models, such as inertia, slow response to market shifts, and failure to capture complete revenue. To overcome the challenge of achieving full coverage of continuous action space and adaptable stability in a highly dynamic environment, we propose a reinforcement learning approach based on Proximal Policy Optimization (PPO) that enables continuous and granular price adjustments to enhance price optimization and maximize profit. The structure of the online store dataset is prepared to include demand elasticity, stock, and competitors' prices. The performance of the proposed PPO managed to set the correct price 91% of the time, earned 12.8 times more in cumulative rewards than DQN, and achieved lower price variability, which makes it a robust candidate for real-time dynamic pricing.
No takes yet. Share an insight, caveat, or question.
Almomani et al. (2025) studied this question.