In this talk, we'll see how computers can understand text by learning about modern concepts in Natural Language Processing such as representation learning and recurrent neural networks. We'll show you the basics behind the tasks of text classification, information extraction, machine translation and dialog systems. We will also cover some Python text processing libraries and how you can start building your next NLP project.

Como podemos abordar um problema com machine learning quando os dados escasseiam? Que abordagens e que cuidados devemos ter?

Learn how to put your best foot forward at this Crush Your Coding Interview by Facebook. This presentation will focus on how to do well in software engineering interviews. It will cover:

• How to walk through an algorithm problem

• How to brainstorm solutions

• How to discuss the problem with your interviewer

• How to write (and test) code on a whiteboard

Machine learning for fraud prevention: from tree based models with feature engineering to deep learning, passing through AutoML.

A Retinopatia Diabética (DR) é a maior causa de cegueira evitável no mundo desenvolvido, afetando mais de 25% dos pacientes diabéticos. Devido aos grandes impactos sociais e económicos da cegueira, a maioria dos países desenvolvidos implementou programas de rastreio para monitorizar a população diabética. Durante o rastreio, imagens do fundo do olho são obtidas e examinadas por oftalmologistas. Nós desenvolvemos um método de Inteligência Artificial que classifica automaticamente a qualidade das imagens, deteta indícios de DR e indica se o paciente deve ser tratado ou não. Nós mostramos que o nosso modelo tem resultados comparáveis a alguns oftalmologistas e obtém os resultados suficientemente bons para ser implementado em programas de rastreio. Este método tem o potencial de reduzir a pressão sobre os oftalmologistas, permitindo que se foquem no tratamento dos pacientes com DR, com o objetivo final de reduzir o número de incidentes de cegueira relacionada com DR.

The Open Food Network and it's different dimensions: the food system, open source, open community, e-commerce platform.

What technology provided the humans leveraged on it.

Blockchain appeared to solve a problem and it’s rapidly changing the world, Bitcoin was the first use case, Libra is the the next big thing, and is aiming to achieve 50M users in record time. But what is Blockchain ? What is behind the scene ? Is C better than Rust for this systems ? I will talk about the technology behind is and what it means for the future of software development.

A aplicação Anda permite aos utilizadores de transportes públicos do Porto usar o telemóvel em vez de um cartão físico para viajar na rede de transportes multimodais. Na data de lançamento desta aplicação, existiam vulnerabilidades de segurança que permitiam aceder a dados de outros utilizadores, incluindo passwords em plain-text. Nesta palestra será explicado o processo de descoberta desta falha. Serão também exploradas quais as ferramentas mais adequadas para realizar este tipo de investigação em aplicações Android (por exemplo: Frida).

Building a product and making the produt grow and evolve differs a lot from what's taught in courses and lectured in classes. More than engineering skills it requires a lot of communication, ownership and leadership skills. In this talk we'll discuss how is to work for an IT company, what drives change, what drives product, what drives engineering and ultimately what should you know before you start your engineering career.

In this talk, I will present an overview of the Malware as a Service (MaaS) environment and show how a major player in this market operates. I will also uncover some threat actors, clients of MaaS providers, that perform cyber attacks on a global scale.