/Fys-Stk4155-Project2

Primary LanguageJupyter Notebook

Ising Model Analysis using linear regregreesion and neural networks

Project doing linear regression, logistic regression and binary classification using a simple neural network on data from the 1D and 2D Ising model

Build with

scikit-image 0.13.1
scikit-learn 0.19.2
python 3.6.2
numpy 1.13.3
matplotlib 2.0.2

Structure of the repo

IsingData

Containg the data for two dimensional Ising Model

Metha

Notebooks from Metha et al

csv

csv files for neural network runs

plots

plots for report

results

Results from linear regression and logistic regression simulations

src

NN_fulldata.py: Program for testing neural network architecture on full dataset from the two dimensional Ising model
NN_partialdata.py: Program for testing neural network architecture on partial dataset from the two dimensional Ising model
NN_lin.py: Program for testing neural network architecture on dataset from the one dimensional Ising model
NN_log.py: Program for testing neural network architecture onpartial dataset
linreg_ising.py: Testing linear regression models on the one dimensional Ising model logreg_main.py: Testing logistic regression on the one dimensional Ising model
fetch_2D_data.py: Handeling the data for 2D ising model
misc: Contians programs used to generate example plots or testing concepts
network: Package

  • NN.py: Class containing neural network methods
  • cls_reg.py: Class containing linear regression methods ols, ridge and lasso; resampling k-fold and bootstrap; statistics MSE and R2
  • logreg.py: Class for doing logistic regression
  • plotparams.py: Parameters for ploting

Authors