Deep Learning Course

9th and 10th June 2022


Modern deep learning applications with Keras

Registration and prices at the bottom.


Until early bird price expires








General Information

Format: In Person
Location: Technopark Winterthur
When: Thursday 9th June, Friday 10th June
Time: 9:00-17:00 (Networking lunch included for both days)
Invited Speakers: Both days an invited speakers will give a talk to inspire and show state of the art applications of deep learning (more information on speakers below).

Course will be held regardless of number of participants.


Beginner/Intermediate Python Programming experience
Basic undergraduate Math knowledge
No previous neural network experience rquired

For whom is the course?

The course is for everyone who is interested in learning the fundamentals of neural networks and deep learning. It is for practitioners who works in industry as for researchers and students in all scientific fields. We will look not only at theoretical concepts but we will have a lot of hands-on session with a lot of ready-to-use code for you to try and adapt to your projects. We will also discuss how to implement models, check them, create custom frameworks in Keras and much more. If you are interested in learning what deep learning is and how to use neural networks this course is for you.

The course is organised in a sequence of blocks. Theory blocks are always followed by hands-on parts where the student will implement in python the thereotical concepts explained in the theoretical part. Each day at 13:00 an invited speaker will inspire the students showing what incredible possibilities deep learning offers.




The days have the following structure:


Theory and Hands-on


Networking Lunch (included in the price)


Invited Speaker


Theory and Hands-on

Number of hours of theory and hands-on

Number of inspirational talks from invited speakers

Number of networking lunches


All topics are covered both in theory and with hands-on sessions. The course is based on the book Applied Deep Learning with TensorFlow 2: Learn to Implement Advanced Deep Learning Techniques with Python by Umberto Michelucci (2nd Edition) (APRESS 2022).

  • General introduction to neural networks and the problem of optimisation

  • Introduction to Keras and TensorFlow 2.X. Fundamentals and advanced topics (custom training loops, custom callback functions, etc.)

  • Neuron: fundamentals components (activation functions, types, etc.). Linear regression and binary classification (logistic regression) performed with one single neuron with hands-application in Keras

  • Feed Forward Neural Networks (multiple neurons)

  • Gradient Descent (and its variation: mini-batch and stochastic gradient descent) and advanced optimisers (Adam, RMSProp, etc.)

  • Hyper-parameter tuning (grid and random search, parameter search on a logarithmic scale, coarse-to-fine search)

  • Regularisation (l1, l2 and dropout)

  • Metric analysis

  • Convolutional neural networks

  • Autoencoders

  • GANs fundamentals

The hands-on sessions are performed with Jupyter notebooks. No installation on student’s laptops will be necessary. All code will be executable in Google Colab.  Only pre-requisites are google chrome and a google account (only if you want to save modifications made to the Jupyter notebooks in a private google drive). A GitHub repository will be available for students with all the codes, material, etc.



Umberto Michelucci

Google ScholarLinkedin

TOELT LLC co-Founder and chief AI Scientist

Teaching Assistant

Arnaud Gucciardi

Early Stage Researcher

  • BSc in pure Maths at the University of Lille, France.
  • 2 years of a Master degree in Computer Science
  • MSc in Bioinformatics in 2021 

Ask for information and register (not binding)

9 + 15 =

Reach Us

You can reach us in several ways. All are good.


Dübendorf, Switzerland

Call Us

+41 (79) 396 7406

Email Us