The code written on the understanding of the paper: "Manifold Regularization: A Geometric Framework for Learning from Labeled and Unlabeled Examples"

1. Dataset The dataset used was the CIFAR-10. The dataset can be accessed through the link given below: https://www.cs.toronto.edu/~kriz/cifar.html

2. Feature Extraction A) Feature_HOG.py : Extracts HOG feature descriptors for the CIFAR dataset B) Feature_intensityvalues.py : Extraction grayscale intensity values for the CIFAR dataset

3. Main Code

A)LapRLS

1. Singleclass_Singleview_permutations_S.py : Implements a supervised classifier for binary classification single-view case.
2. Singleclass_Singleview_permutations_SS.py : Implements a Semi-supervised classifier for binary classification single-view case.
3. Singleclass_Multiview_permutations_S.py : Implements a supervised classifier for binary classification multi-view case.
4. Singleclass_Multiview_permutations_SS.py : Implements a Semi-supervised classifier for binary classification multi-view case.
5. Multiclass_Singleview_permutations_S.py : Implements a supervised classifier for multi class and single-view case.
6. Multiclass_Singleview_permutations_SS.py : Implements a Semi-supervised classifier for multi class and single-view case.
7. Multiclass_Multiview_permutations_S.py : Implements a supervised classifier for multi class and Multi-view case.
8. Multiclass_Multiview_permutations_SS.py : Implements a Semi-supervised classifier for multi class and Multi-view case. The codes were run on Anaconda (Spyder) – Python 2.7.

B) LapSVM

1. Laplacian Support Vector Machine Single-view single class – LAPSVM_SV_SC_SS.py
2. Laplacian Support Vector Machine Single-view multi-class – LAPSVM_SV_MC_SS.py
3. Laplacian Support Vector Machine Multi-view single class - LAPSVM_MV_SC_SS.py
4. Laplacian Support Vector Machine Multi-view multi-class – LAPSVM_MV_MC_SS.py For supervised learning set the gamma_I (gamma_I1, gamma_I2) parameters in all the scripts as zero Python version 3.6 is used for implementing and running all the scripts. Install CVXOPT V 1.1.9 library that is compatible with the system and python version used for running the scripts.