Skip to content

Latest commit

 

History

History

Effective_dimension_plots

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
data
data
 
 
generate_data
generate_data
 
 
plot_fig
plot_fig
 
 
.DS_Store
.DS_Store
 
 
README.md
README.md
 
 

README.md

Code to reproduce the effective dimension plot

In this folder, there are 3 subfolders each containing the following:

  • data: raw data of the normalised Fisher information matrix for the classical neural network, the easy quantum model and the quantum neural network. All models have an input size of 4, output size of 2 and number of parameters d = 40.

  • generate data: in this folder, there are 3 separate code files to reproduce the Fisher information matrices for all 3 models discussed above. This is required for the calculation of the effective dimension. These code files require functions from the effective_dimension package.

  • plot_fig: this folder contains a code file to convert the Fisher information matrices to the effective dimension over different values of data (n) and plot these results. There is also a .txt file containing the raw source data of the effective dimension values.