This work was supported by the research project RASTA: Realtá Aumentata e Story-Telling Automatizzato per la valorizzazione di Beni Culturali ed Itinerari; funded by the Italian Ministry of University and Research (MUR) under the PON Project ARS01_00540.
The dataset folder contains the methods used for creating the dataset utilized
in training and testing the models. Additionally, it includes a notebook for
analyzing the dataset, allowing for a detailed exploration of its characteristics
and insights.
The dqn folder contains the implementation of the DQN (Deep Q-Network) model.
It includes the following files:
Agent_DQN.py: Contains the implementation of the DQN agent.Environment_DQN.py: Contains the implementation of the environment used for the model.Dataset_DQN.py: Implements methods for managing the dataset.Settings_DQN.py: Includes the configuration settings for the DQN model.
To run the experiment, use the notebook dqn.ipynb, which provides a complete
process for training and testing the DQN model.
Below is an image representation of the DQN model:
The profile_models folder contains the implementation of the baseline profile
models used for comparison against our model. This folder includes all the necessary
files to run the experiments, along with a notebook that provides a complete
workflow for training and testing these models.