Introduction to Deep Learning
In this talk, I will introduce the core concepts for the understanding of neural networks. I will start from very basic concepts that everyone should be familiar with, such as numbers and basic operators (addition, subtraction and multiplication), and systematically guide the audience towards every essential concept for the understanding of how neural networks function. Using a simple example based on a popular educational cartoon, we will start by developing a simple linear model, in order to illustrate how to define a model and optimizer its parameters. Then, we show how more complex models can be built (e.g. Multilayer Perceptrons), and describe the key technologies that enable, such models to be built and trained (e.g. GPUs, Computational graphs).
Translation Quality Estimation
In this talk, I will present a tutorial on translation quality estimation, a task of growing importance in NLP, due to its potential to reduce post-editing human effort in disruptive ways. In particular, I will present Unbabel's quality estimation system, where we achieve remarkable improvements by exploiting synergies between the related tasks of word-level quality estimation and automatic post-editing. First, we stack a new, carefully engineered, neural model into a rich feature-based word-level quality estimation system. Then, we use the output of an automatic post-editing system as an extra feature, obtaining a new state of the art for word-level and sentence-level quality estimation. I will end with some thoughts about future work in this area.
Imitation learning for structured prediction in natural language processing
In this talk we will see how to use imitation learning to improve incremental structured prediction models with applications to natural language understanding and generation.
Introduction to AI
In this workshop we will introduce you to basic concepts of Artificial Intelligence, using a very hands-on approach. Being an introductory workshop we will cover a broad range of subjects from different sub-fields of AI, like reactive agents (or bots), search algorithms, or evolutionary computation. Using games as an example environment, we will implement some of the algorithms mentioned, providing a better (and hopefully more enjoyable) way to learn about them.Instructions
In order to participate in Tiago Baptista's workshop you need Python 3.5 or Python 3.6 installed (as well as pip3). You also need the following things:
1.IDE (ideally PyCharm)
2.pyafai — which can be installed with "pip install pyafai".
Make sure you install pyafai with pip3 for Python 3 (you might need to run "pip3 install pyafai" on some systems).
Markov Chains for Text Generation Tutorial
In this tutorial we will go through the basics of markov models and implement them to achieve a toy use case (generate text).
We will learn what a markov model is, why they are useful to model certain processes, and how to implement one and train it. With one implemented we will train it on existing test data (we will bring a sample of text to train the model on, but feel free to bring your own), and generate text.Instructions
In order to participate in Rodrigo Gomes's workshop, you need Python 3 (and pip3) installed.
The list of packages you need for the workshop will be revealed later or during the workshop.
Deep Learning with Tensorflow
This workshop will be an introduction to deep learning. We will discuss the advantages and disadvantages of deep learning (when to use, when not to use), understand how neural networks are actually trained, and get our hands dirty with Tensorflow in order to solve tasks like digit recognition.Instructions
In order to take part in Miguel's workshop, you will need Python 3 (and pip 3) and follow the instructions here:
A gentle introduction to recommender systems — content-based systems v.s. collaborative filtering, dimensionality reduction, similarity metrics.Instructions
The workshop is language-agnostic and the participants will develop recommender systems algorithms in their preferred programming language. Here is a paper about the workshop.
Visualizing your ML Models
Pedro Fonseca & Samuel Hopkin
In this workshop you will learn some intuitions about how different machine learning algorithms make decisions, by training different models and observing the resulting decision boundaries. It will be a highly visual approach into the different concepts of machine learning, and also of overfitting, and the tradeoff between variance and bias.Instructions
The instructions for Pedro & Sam's workshop can be found and followed here.