Dynamický pricingový model s využitím predikčních modelů

The thesis is describing an algorithm of dynamic pricing model for optimal price setting in e-commerce, specifically for internet search engine. Pricing in e-commerce is very different and is based on a specific line of business, since the price is not dependent on products' prices but it is driven bz customers' behaviour and companies' revenue comes directly from exit clicks made on their website. The aim of the thesis is to introduce a complex topic of dynamic pricing, how to approach it and how to optimize revenue of internet search engines with respecting their clients' COS (Cost of Sales). Pricing model is based on forecasting models and the thesis is comparing results from SARIMA and Prophet, which was introduced by Meta in February 2017, with different models' parameters. Apart from finding the most optimal configuration of the model, the thesis is also advocating in favor of using forecasting models and machine-learning in pricing which demostrated better results instead of relying on employees and their intuition.