/IAII_II

Feedforward Neural Network using Backpropagation for solving pattern classification problems.

Primary LanguagePython

Feedforward Neural Network using Backpropagation.

Implementation of a feedforward Neural Network using backpropagation for solving pattern classification problems.

Requirements:

  • Python 3.x
  • Numpy
  • Matplotlib
  • Pandas

Running:

$ python excercise2.py

Will perform all networks (2 to 10 neurons) for all datasets.

$ python excercise2_bestset.py

Will perform all networks (2 to 10 neurons) for only for the best dataset.

$ python excercise3_p1.py

Will run prediction for binary classification. As in Excercise 3 Part 1.

$ python excercise3_p2.py

Will run prediction for multiclass classification. As in Excercise 3 Part 2.

Other:

  • Graphics and results are stored in results/ folder.
  • A complete description of this proyect can be found in the docs folder.

Authors:

Last Updated: 04/03/2017