Active_learn

The 64x64 greyscale images used for testing and training are in the "dataset" folder. The "flatten_images.m" file takes these images and creates a "dataset.mat" file with training/test data/labels.

• The "active_learn.m" file executes active learning with AdaBoost, which can use decision trees, neural networks, or convolutional neural networks as weak learners.
  • The "train_boosted_dt.m" file trains AdaBoost, and "test_boosted_dt.m" tests it.
    • The "neural_train.m" file trains a neural network, and "predict_net.m" gives the network's prediction.
      • The "cnn.m" file trains a convolutional neural network.
  
  The dataset in "dataset.mat" format is also available

here