Deep Learning

5.0 - - 2 reviews

You'll be able to :

Understand and implement state-of-the-art deep learning solutions;

Use deep learning algortihms in realistic projects;

Understand the strength and the limitations of deep learning.

Requirements :

Intermediate programming skills in Python, and experience with Numpy and Pandas;

Have general knowledge about supervised learning;

Laptop with minimal computing power: RAM>4GB, GPU.

Course features

Lecturer	Valerio Rossetti
Starting	03/Novembre/2020
Ending	12/Decembre/2020
Training Sessions	Tuesday and Thursday, 7pm-9pm Saturday, 8.30am-12.30am
Location	Ecole-club Migros Genève
# hours	40
Max Participants	16
Language	English
Full Price Registration	1600 CHF

Description

With this course, you will be able to understand and implement state-of-the-art solutions in deep learning, such as convolutional neural networks, recurrent neural networks, and transfer learning. The course will gradually bring you from the simple perceptron model to powerful algorithms, such as the LSTM models, which are used every day in our phones for voice recognition.

The course includes two full-immersion Saturdays of 8 hours. During these days, you will be able to work on a realistic project in sectors such as healthcare, engineering or financial trading.

This course is tailored to professionals who have already a basic understanding of machine learning and that would like to learn to use advanced techniques. It is also suitable for academic researchers who are looking for career opportunities in the private sector.

How it works

Registration

You can pay immediately if you think you have the right level for the course. You will receive a confirmation message. If you are not sure, you can ask for an assessment. We will meet you and discuss with you what are your goals and what is your level. At the end of the assessment, we will tell you what we think is the right choice for you.

Connect with amazing peers

During the course, you will interact with participants coming from industry and academia. Coding together and making small hackathons is the best way to learn!

Certificate

After completion of a project that you will show to experts in the industry, we will provide you with a certificate of completion that shows that you commited until the very end.

Ask for information

Content

Module 1:

Brief introduction to Artificial Neural Networks

From linear regretion to Multi-Layer Perceptron (MLP)

MLP in Scikit Learn and Keras

Strengths and limitations of MLP algorithms

Module 2:

Deep Neural Networks

Deep Feedforward Networks: theory and implementation in Keras

Regularization, tuning, optimizers and best practises

Autoencoders: theory and applications

Module 3:

Convolutional Neural Networks

Convolutional Neural Networks: theory and implementation in Keras

Transfer Learning

Introduction to object detection

Full-Immersion 1

Build an real-world deep learning solution.

Module 4:

Recurrent Neural Networks

Sequence Modelling

Long-short Term Memory Nets

Module 5:

Deep Learning and Natural Language Processing

Introduction to Natural Language Processing (NLP)

NLP with spaCy

Module 6:

Reinforcement Learning

Brief introduction to Reinforcement Learning

Full-Immersion 2

Build an real-world deep learning solution.

Reviews

- Admir Greljo - CERN

Deep Learning - December 2018

Deep learning is used more and more in theoretical particle physics, my main area of expertise. I enrolled in this course with the goal to be able to follow and contribute to the literature. The course provided a solid theoretical foundation of the subject, as well as, an immense number of practical example, which I could straightforwardly apply to my research.

Following the completion of the course, I have no difficulties understanding the ML seminars given at CERN, and I started using these techniques in my collider projects. Therefore, I am delighted to give the highest possible recommendation for this course. This positive experience already enriched my academic career. Thanks to SamurAI.

- Sviatoslav Zimine - Metafused

Deep Learning - December 2018

I was happy to attend the course on Deep Learning offered by Samurai. The most important DL models are covered with a focus on model code implementation examples. The contact with the lecturer is permanent for any questions raised.

To get full benefits from this course, expect to invest a fair effort in learning as well as in the immersion project. Many thanks to Fabio and the team for rolling-out this program.

FAQs

Does deep learning have real-life applications?

Deep learning has many applications in the real world. Among them, there is text translation, speech recognition, and computer vision, to cite a few. One particular application that has the power to reshape our future is cancer diagnosis through the help of deep convolutional neural networks when applied to medical images.

Can I just enroll in a few modules, instead of taking the full course?

We, typically, don't accept people for only a few modules, as we think that a full grasp of the course is necessary to be able to understand which techniques to apply in the real world. Moreover, by only taking a selective number of modules, we will not be able to deliver you the certificate at the end.

Do I need special hardware for deep learning?

Deep learning techniques rely heavily on what is called "tensorial calculations" that can be parallelized through a GPU. It will therefore be advisable to bring a laptop with a GPU.

How do I know if this is the right course for me?

This course is a unique opportunity to understand what people exactly mean by deep learning and how it is applied in fields such as healthcare, autonomous transportation, robotics or financial trading. If you are interested in such topics and advanced analytics, then please see the enrollment prerequisites.