Neural Networks (2018-2019)

Older editions: 2012 – 2013/2014 – 2016/2017 – 2017/2018


This page contains the material for the lab sessions of the Neural Networks course taught by Prof. Aurelio Uncini for the M.Sc. in Artificial Intelligence and Robotics. For info about the course refer to the original web page.


Lab sessions require a working installation of Python 3.6 (or higher) with (at least) Jupyter, NumPymatplotlibscikit-learn, and TensorFlow.

  1. If you are starting from scratch, installing the Anaconda platform is the recommended method to have a working installation with all prerequisites (except TensorFlow).
  2. If you prefer not to install packages, you can work from the web on the Google Colab service.

List of lab sessions

Lab session 1 (Python and NumPy) [Colab / GitHub]: get familiarity with Python, NumPy, and working with notebooks.

Lab session 2 (Linear regression) [Colab / GitHub]: implement a linear regression and a polynomial regression directly in NumPy.

Lab session 3 (TensorFlow eager execution) [Colab / GitHub]: basics of TensorFlow eager execution, simple classifier on the Iris dataset.

Lab session 4 ( and convolutional neural networks) [Colab / GitHub]: CNN classifier on MNIST, loading data with