Homepage of Armando Vieira

Professor, Entrepreneur, Data Analyst and Scientist

  • My researh

    Armando Vieira

    After my PhD at University of Coimbra, I spend a year in Grenoble, France, as a Postdoc. Then I decided to do something new.
    From 1998 to 2000 I worked on the Portuguese Nuclear Reactor at Sacavém, studying neutron diffusion and radiation protection.

    Since 1998 I fell in love with Artificial Neural Networks (ANN) and Machine Learning. I applied these techniques to any problems that I could put my hands on: from physics to economics or even deep oil discovery. We were able to successful apply ANN to several Rutherfod  BackScattering (RBS) spectra analysis of thin films and revolutionize the field of real-time RBS.

    In 2003, I was able to propose a new algorithm to train Neural Networks for classification problems and apply it to complex tasks, namely financial distress prediction of private French companies.

    In 2004 I took a Sabbatical leave at Los Alamos National Laboratory to engage in the study of Complex Network Analysis.  A topic that I eventually resume early in 2011 and in which I have a keen interest, especially in its applications to recommendation systems and correlated credit events (dominó bankruptcies). Early 2011 Tiago and I were the first portuguese team to be accepted at Seedcamp London, the biggest entrepreneurship meeting in Europe.

  • My Passions

    Data analysis, algorithms, machine learning, social networks, complexity.

  • My Education

    I had the privilege to take my degree of Physics, in 1990, at one of the best Portuguese engineering schools, Instituto Superior Técnico, in the beautiful city of Lisbon.
    In 1993 I took the master degree in Plasma Physics to study the hot plasma of the ISTOK tokamak using function parametrization. I moved to Coimbra in 1993 for my PhD in Theoretical Physics. Though I didn’t love the city, I got fascinated by the personality and character of Professor Carlos Fiolhais. The thesis was on fission of metallic clusters – building blocks of nanotechnology.

  • Teaching

    Beside being a full professor of physics at ISEP, I lectured on several international post-graduated degrees, like master in Bioinformatics in University Pompeu Fabra in Barcelona and at a MBA on real estate in ISEG, Lisbon.
  • Some projects

  • See this demo on Latent Dirichelet Allocation for a dating web site.

    04
    FEB
    2016

    Open Source Libraries for Deep Learning

    There are many open source libraries available: Caffe TensorFlow, Theano, and Torch.   Caffe is one of the first deep learning toolkit, started in late 2013, mainly used for convolutional neural network and still actively used. However, it doesn’t support recurrent networks...
    27
    JAN
    2016

    Symbols and meaning

    I just come across a recent publication of Nate Soares by the Miri institute https://intelligence.org/files/RealisticWorldModels.pdf concerning a theoretical framework on artifical cognitive agents building models of the world. The author refers to the Solomonof induction problem...
    19
    JAN
    2016

    Knowledge Representation and Visualisation in Graphs using Convolutional Neural Networks

    My newest paper is “Knowledge Representation and Visualisation in Graphs using Convolutional Neural Networks”. Here is the abstract: “Knowledge Graphs (KG) constitute a flexible representation of complex relationships between entities particularly useful for...
    14
    JAN
    2016

    Fresh perspective on the symbol grounding problem

    A striking feature of the human brain is the ability to associate abstract concepts with the sensory input signals, such as visual and audio. As a result of this multimodal association, a concept can be located and translated from a representation of one modality (visual) to...
    10
    JAN
    2016

    Neural Networks for symbolic deductive reasoning?

    The goal of Artificial General Intelligence (AGI) is to create computer systems with human-like intelligence. AGI systems should be able to reason and learn from experience by interacting with the environment in a mostly unsupervised way. To build intelligent machines it is...
    21
    DEC
    2015

    My perspective on Tensorflow

    Tensorflow is a framework for deep learning recently released by Google. It is based on a set of Theano libraries (heavily relying on tensors – thus the name of the project). But it’s much more than Theano. Together with the symbolic manipulation of variables, much in...
    30
    NOV
    2015

    Digging for gold: how machine learning can disrupt the IT talent acquisition market

    Talent acquisition and retention is the single most determinant factor of success in IT companies. It’s also the most cumbersome, unreliable and stupid process I ever experienced. Being on both sides of the fence, I got astonished by the lack of preparation of candidates applying...
    20
    OCT
    2015

    A new dating algorithm (part 2/4)

    Knowing me knowing you Topic modeling is essentially a statistical technique to uncover the hidden subjects, or topics, that occur in a collection of documents.  The method is completely unsupervised and the learning is performed based on subtle co-occurrences of words in a...
    06
    OCT
    2015

    Seasonality effects in cars sales using machine learning

    Seasonality effects on second hand cars sales from Armando Vieira