Theoretical and practical introduction to neural networks and their applications. The theoretical part will describe how Frank Rosenblatt’s Perceptron model has evolved into today’s Deep Learning networks. Theoretical concepts such as loss function, the gradient descent method, metrics and the confusion matrix will be reviewed. We will demonstrate how to get a dataset ready for an experiment and how to choose a network to solve a specific task: classification networks, object detection networks and semantic segmentation networks.
In the second part, the knowledge gained in the theoretical part will be put into practice using the Google Colab environment. We will show how to do experiments harnessing a Jupyter Notebook. At the end of the session, the Jupyter Notebook will be available for attendees to use to continue experimenting.
Addressed to:
-People who are interested in the potential of Deep Learning and would like to conduct an experiment. Knowledge of programming (preferably in Python) and linear algebra is recommended.
