/markov_simulation

Simulation of customer behaviour in a supermarket based on Markov Chain.

Primary LanguageJupyter NotebookMIT LicenseMIT

Markov Simulation

This project simulates a customer behavior in a supermarket based on Markov Chain. The project has been developed in collaboration with @ai-aksoyoglu, @AlphanAksoyoglu and @pavrmk and involves the following steps:

1. Exploring the data (includes pandas wrangling)

See the notebook in /data_exploration

2. Calculating transition probabilities (a 5x5 matrix)

See the notebooks in /transition_probas

3. Creating a customer class and implementing MCMC for one customer

See the notebook in /mcmc_for_one_customer

4. Creating a supermarket class that simulates multiple customers and adding it together with the customer class to a module

See the final module supermarket_simulation.py

5. Filling the shelves of a supermarket image and visualizing the supermarket

See the notebook in /supermarket_visualization