Past Meetup

**FULLY BOOKED** 2 DAY WORKSHOP: FUNDAMENTALS OF PRACTICAL DEEP LEARNING

This Meetup is past

39 people went

Location image of event venue

Details

IMPORTANT: Places are only guaranteed if you registered via the workshop registration page (http://www.eventbrite.com/e/2-day-workshop-fundamentals-of-practical-deep-learning-tickets-19950545619). Click on the link below to register and reserve your seat. > http://www.eventbrite.com/e/2-day-workshop-fundamentals-of-practical-deep-learning-tickets-19950545619 (http://www.eventbrite.com/e/2-day-workshop-fundamentals-of-practical-deep-learning-tickets-19950545619)

NVIDIA in partnership with Persontyle are excited to announce this amazing 2-day hands-on deep learning workshop for developers, data scientists, researchers and quantitative analysts. Great opportunity to learn and practically get started in using deep learning methods and frameworks.

Overview:

Machine Learning is among the most important developments in the history of computing. Deep learning is one of the fastest growing areas of machine learning and a hot topic in both academia and industry. It has dramatically improved the state-of-the-art in areas such as speech recognition, computer vision, predicting the activity of drug molecules, and many other machine learning tasks. The basic idea of deep learning is to automatically learn to represent data in multiple layers of increasing abstraction, thus helping to discover intricate structure in large datasets.

This workshop will cover the fundamentals of deep learning, starting from definitions of neural networks and their learning criteria, specific aspects of deep networks, optimization, regularization, convolutional structures for image data, recurrent neural networks for sequence modelling, as well as supervised and unsupervised learning.

Target Audience:

This course is targeted for developers, data scientists or data analysts that have a basic knowledge of machine learning.

Prerequisites:
Experience on programming, basic knowledge of calculus, linear algebra, and probability theory. Attendees are expected to bring their own laptops for the hands-on practical work.

Registration fee:
Corporate attendees: £ 400 + VAT (http://www.eventbrite.com/e/2-day-workshop-fundamentals-of-practical-deep-learning-tickets-19950545619), including course material as well as breakfast, lunch and coffee service on both days.

Special Discount for Academics and Student attendees: please contact us for a reduced rate.

A limited number of seats are available. Seats are filled on a first come, first serve basis.

* Workshop Agenda:

DAY 1

9:30-10:00 Registration and Breakfast

10:00-10:15 Welcome

10:15-11:15 Theory: Introduction to Deep Learning

• Example applications show the impact of deep learning in different fields. Basic concepts and definitions are introduced with a simple example network.

11:15- 11:30 Coffee Break

11:30-12:45 Theory: Optimization and Initialization

• Training deep networks happens typically by minimization of a training criterion. This optimization problem is very high-dimensional, non-convex, and typically non-constrained. The signals propagating in the network tend to decay or explode exponentially w.r.t. the depth of the network, which has to be taken into account.

12:45-13:30 Lunch

13:30-14:30 Theory: Regularization

• Deep networks have great expressive power, which means that good performance on the training set does not guarantee that the results can be generalized to new cases. There are a number of regularization methods that are used to alleviate this problem known as overfitting.

14:30-15:00 Hands-on lab: Introduction to NVIDIA software for Deep Learning

● Benefits of parallel computations

15:00-15:15 Coffee Break

15:15-17:00 Hands-on lab: Caffe framework

How to:

• Build and train a convolutional neural network for classifying images.

• Evaluate the classification performance under different training parameter configurations.

• Modify the network configuration to improve classification performance.

• Visualize the features that a trained network has learned.

• Classify new test images using a trained network.

• Training and classifying with a subset of the ImageNet dataset.

• 17:00-17:15 Summary of the learning goals and day 1 close

DAY 2

9:00-9:30 Breakfast

9:30-9:45 Recap of Day 1

9:45-10:45 Theory: Convolutional Networks

• Convolutional networks have revolutionized the field of computer vision by dramatically improving the state-of-the-art performance in recognition and other tasks. Convolutional networks are deep models that use the known spatial structure of pixels and invariance to small translations as prior information.

10:45- 11:00 Coffee Break

11:00-12:00 Theory: Recurrent Neural Networks

• Recurrent neural networks (RNN and variants such as the LSTM) are useful in case of time series analysis. The best speech recognition and machine translation models are based on them. RNNs can be seen as networks where consecutive time points are connected, thus forming an infinitely deep network unfolded in time.

12:00-12:45 Lunch

12:45-13:45 Theory: Unsupervised Learning

• Many of the recent successes of deep learning are based on reproducing the target outputs provided by human experts in the training data. A future trend in deep learning is towards unsupervised learning, where the target outputs are not available, but the learning is based on analysis of the input data itself. Unsupervised learning can also support supervised tasks such as classification, an idea used in Ladder networks, leading into state-of-the-art performance.

13:45-14:45 Hands-on lab: Torch framework

• Introduction & History

• Torch core features

• Why use Torch?

• Torch Community and support.

• The Cheatsheet

• Lua (JIT) and LuaRocks

• Torch’s universal data structure

• Tensors

Creating a LeNet network
Criterion: Defining a loss function
Using dataloaders to load 50,000 CIFAR-10 (3x32x32) images
Load and normalize data
Define Neural Network
Define Loss function
Train network on training data
Test network on test data.

14:45-15:15 Coffee Break

15:15-16:45 Hands-on lab: Theano framework

By Pyry Takala / PhD Student, Aalto University

16:45-17:30 Summary of the workshop and closure

18:00 Social and Networking

*(exercises completed and schedule can be changed, notice will be provided)

Workshop Instructor and Lab Trainers:

• Dr. Tapani Raiko (https://users.ics.aalto.fi/praiko/), Assistant Professor, Aalto University, Finland. The theoretical sessions will be delivered by Dr. Raiko.

• Pyry Takala, PhD student, Aalto University, Finland.

• Alison B. Lowndes, Deep Learning Solution Architect, NVIDIA, UK.

Please visit the workshop page to register. (http://www.eventbrite.com/e/2-day-workshop-fundamentals-of-practical-deep-learning-tickets-19950545619)