Node2Vec & Metapath2Vec
Graphs
Graphs are used to represent real systems and the interaction among their components. Protein interactions, neurons organization/communication, human interactions and societal behavior are some examples [1].
Graphs are constituted of vertices and edges (Figure 1). And they are the object of study of Graph Theory, which uses a set of formalities (e.g.: nomenclature) indispensable for their study [1].
Although graphs are a convenient way of representing data, there are several reasons to consider graph embeddings:
- To perform machine learning in graphs directly is restrictive. Vector spaces have a richer toolset of of approaches [3].
- Compressed representation of graphs. While an…
BERT, RoBERTa & GPT-3
Introduction
Unlike in Computer Vision, in Natural Language Processing (NLP), pre-trained models only became widely available recently. In part due to the scarcity of text datasets, NLP had a slow progression until the release of Bidirectional Encoder Representations from Transformers (BERT) in 2018 [1].
BERT was published by Jacob Devlin and his colleagues at Google AI Language [1].
There are two pre-trained models available of BERT: base and large. Although both use the same architecture, the first contains 110 million parameters, while the second 345 million. Being able to use a pre-trained model and fine-tune it to a different task means…
PCA & T-SNE
Introduction
Dimensionality reduction aims to preserve as much information as possible from higher dimensional vectors. Principal Component Analysis (PCA) [1] and T-Distributed Stochastic Neighbouring Entities (T-SNE) [2] are two of the most commonly used approaches. While the first is considered having a mathematical approach to the problem, the second uses a statistical one.
The main goal of PCA resides in preserving the vectorial components with higher variability across the data, while discarding the ones adding less information. This decomposition can be achieved in two different ways. One is by decomposing the data covariance matrix in its eigenvalues. …
Erasmus Paris, França
História e Reconhecimento
Iniciada a construção após o final da I Guerra Mundial, a Cité Internationale Universitaire de Paris acolhe, desde então, estudantes de múltiplos países. Neste momento, cerca de 12 000 estudantes e investigadores encontram-se distribuídos pelas 40 residências que compõem este gigantesco espaço parisiense, idealizado por André Honnorat.
De destacar, a candidatura ao prémio Nobel da Paz (2014) ganho pela paquistanesa Malala Yousafzai.
Localização
Edificada numa das zonas mais centrais de Paris, na conhecida Ilê-de-France. Tendo como consequência imediata, a disponibilidade de uma panóplia de diferentes meios de transporte:
Serviço partilhado de bicicletas (tradicionais de cidade) e carros elétricos, a.k.a Vélib e Autolib…
A step-by-step tutorial that will guide you through the complete development process. It also includes prototype testing. A success!
Abstract
As a result of an aging population and a progressively greedy will for better healthcare, new medical devices have arisen. Some focus on prevention, others on treating. The prototype developed here focuses on the acquisition of the heart’s electrical activity with a simple electrical circuit embedded in an Arduino microcontroller, with the possibility of early heart disease detection. An effective and affordable solution was proposed and dozens of trials were performed with the device. These pointed to a high rate of detection of R, S, and T-peaks (~100%, 78.3%, and 72.5%, respectively) with an accurate heart rate measurement. …
