Eingabe löschen

Kopfbereich

Schnellnavigation

Hauptnavigation

New course: Deep Learning Fundamentals

Over the last years, Deep Learning has emerged as the new Swiss army knife of big data analytics. On October 28, our course «Deep Learning Fundamentals» will start. The course is offered in collaboration with Umberto Michelucci, Head of «AI Center of Excellence» at Helsana.

In this interview Umberto Michelucci, Head of «AI Center of Excellence» at Helsana, explains what Deep Learning is and what you can learn in this course.

  1. Can you briefly explain what deep learning is?
    Umberto: «One talks about deep learning when neural networks with many layers are used. Nowadays one speaks about deep learning when dealing with networks with millions, if not billions of parameters.  Typically, the datasets used are also very large. Training such large networks brings a completely new set of challenges that are not encountered in more classical machine learning (like decision trees, regression, etc.). This course goal is to prepare the students to those new challenges and give them a sound foundations to start using deep learning in their projects.»
  2. Can you describe some application areas of deep learning?
    Umberto: «The impressive thing about deep learning is that it has been applied successfully in many areas: computer vision, language understanding, language translation, reinforcement learning and many more. Its range of application is almost infinite, if enough data is available!»
  3. What are the advantages of having deep learning knowledge?
    Umberto: «Understanding what it means to train deep neural networks, will give the students a decisive advantage. Working with large networks requires a completely different mindset that must be learned. Getting results requires experience and theoretical knowledge. If the student wants to start working on more advanced machine learning projects, deep learning is a necessary step.»
  4. The internet is full of courses on deep learning. What is different about your course? What will participants learn?
    Umberto: «The problem with almost all internet courses, is that they promise to make you an expert in a couple of weeks (or even a few days). This is simply lying to students. There are simply too many things to learn. A more realistic timeframe to become experts is probably 2 to 3 years at least. This course aims to give a solid foundation on deep learning to students. Its aim is not to make you an expert, because this would be impossible, but it will give you the necessary knowledge to be able to become an expert. It will kick-start your deep learning journey. After this course you will be able to take the next steps and start working on more complex problems and slowly become and expert.»
  5. Who is the course for?
    Umberto: «This course is for everyone who has a basic to intermediate Python programming experience and is comfortable with algebra and calculus. We will not calculate complicated integrals, but you should know what a derivative or gradient is and should be able to understand matrix multiplication. A background in data science is of course a great advantage. It is important to remember that all the things that are good practices in machine learning (and statistics) are also important in deep learning. Even in deep learning you will have to clean and understand your data, understand the problem you want to solve, deal with missing values, and so on. One thing is of fundamental importance: you should be curious and willing to learn something new and have fun. Because working with deep learning is a lot of fun!»

The course starts on October 28 and takes place in the evenings. Application deadline is October 14. More information and registration: Deep Learning Fundamentals

Learn more about the wide range of continuing education courses offered by the Institute of Applied Simulation in the field of Computational Science and Artificial Intelligence: www.zhaw.ch/ias/weiterbildung