Entanglement Estimation

This repository includes all the code used in the completetion of the BSc project. See the report (BSc_Project_Report.pdf) for details.

Prerequisites

You will need the following python packages to run the code in this repository.

Numpy
Scipy
Pandas 
Matplotlib
Keras
Tensorflow

List of Files

• data_generation.py - Generates simulated experimental data used for training neural networks and monte carlo methods. Data are saved as pandas dataframe and can be export to csv.
• prob_integral.py - Perform the integrations for the Bayesian methods.
• utilities.py - Helper functions.
• hyper_opt_exact.py - Hyperparameter optimisations where errors are calculated using the Bayesian methods.
• hyper_opt_monte_carlo.py - Hyperparameter optimisations where errors are calculated using the testing data generated with data_generation.py.
• plot_error_exact.py - Summarises the root mean squared error (RMSE) for SDI, BME, and the maximum possible RMSE.
• plot_error_monte_carlo.py - Summarises the root mean squared error (RMSE) for SDI using data generated with data_generation.py.
• BSc_Project_Report.pdf - Includes everything you need to know about the project.
• dataset/10k10k_m5_10_20_40.csv - Dataset for N (no. of measurement) = 5, 10, 20 and 40. For each N, there are 10,000 training data and 10,000 testing data.
• dataset/10k10k_m10_100_1000.csv - Dataset for N (no. of measurement) = 10, 100 and 1000. For each N, there are 10,000 training data and 10,000 testing data.
• dataset/10k2k_m5_to_m1000.csv - Dataset for N (no. of measurement) = 5, 10, 50, 100, 500 and 1000. For each N, there are 10,000 training data and 10,000 testing data.