AdaBoost-CNN: An Adaptive Boosting algorithm for Convolutional Neural Networks to classify Multi-Class Imbalanced datasets using Transfer Learning
For All use of the data, please 'cite' the following:
'Aboozar Taherkhani, Georgina Cosma, T.M McGinnity. AdaBoost-CNN: An Adaptive Boosting algorithm for Convolutional Neural Networks to classify Multi-Class Imbalanced datasets using Transfer Learning, Neurocomputing'
The paper is avilabel at: https://doi.org/10.1016/j.neucom.2020.03.064
-To run the AdaBoost-CNN and a single CNN on the synthetic data run ‘test2_CNN.py’.
-The code will generate random synthetic data, and it trains and tests the two methods (AdaBoost-CNN and a single CNN) on this data.
-The variable ‘n_estimators’ before ‘Ada_CNN’ has been set to 10. It determines the number of estimators in the AdaBoost-CNN.
-The variable ‘epoch’ sets the number of training epochs of a CNN estimator in the AdaBoost-CNN.
This code is tested on Python 3.5.4 |Continuum Analytics, Inc.| (default, Aug 14 2017, 13:41:13) [MSC v.1900 64 bit (AMD64)]. It impalement CNN by Keras Using TensorFlow backend.