Deep Learning

Full repository for the Unimi Deep Learning course.

1-Introduction to Python

Basic python: functions and classes

2-Libraries for Data Science

Introduction to: Numpy, Matplotlib, Pandas and TensorFlow

3-Custom models

Sequential models creation from basic linear algebra

4-Regression and Classification Models

Use of keras to perform a regression on a set of points and a classification between images of clothes

5-Hyperparameters tuning

Use of Hyperopt to perform a hyperparameter search.

6-Callbacks and RNNs

Use of callbacks to stop the training process and avoid overfitting.

Use of LSTM to forecast daily temperature.

7-Image Recognition

CNNs, classification and localization.

8-Data augmentation and transfer learning.

Use of data augmentation to improve performances.

Transfer learning from a base model.

9-Generative models

Use of a Generative Adversarial Network to create MNIST numbers from random noise.