Introduction to Deep Learning

Welcome to the Deep Learning course! In our introductory approach, we will use Keras’ functional API, a Deep Learning “building blocks” composed of the TensorFlow framework. Additionally, we will delve into the mathematical underpinnings of artificial neural networks to better understand how to construct and enhance our models.

Contents

The whole course is divided into two modules. Their contents may change throughout the classes to meet the public's needs.

Module I:

• Deep Learning workflow;
• Basic math of neural network modeling and training;
• Tasks: classification, regression, clusterization, (fine-tuning);
• Dense, Convolutional, and Recurrent layers;
• Dataset types: dataframe, images, time-series, counts table.

Module II:

• Object oriented programming in Python (OOP);
• Tensorflow/PyTorch OOP;
• Generative models: GAN, VAE;
• Graph Neural Network (GNN);
• Large Language Models (LLMs);
• Model fine-tuning;
• Model interpretability.

Required Python modules

Please make sure to install the following Python modules for this course (pip install module-name):

• numpy
• scipy
• pandas
• matplotlib
• tensorflow

You can install the dependencies using a conda recipe

conda env create -f environment.yml

References

• Deep Learning with Python;
• Neural Networks and Deep Learning;
• Deep Learning Book;
• Neural Networks - 3Blue1Brown

About the author

• GitHub
• ORCID