Anibal Solon
Brazilian, 26 years old
Educational Experience
MSc in Computer Science
Pontifícia Universidade Catolica do Rio Grande do Sul / Brazil – 2014-2016
Thesis: Deep Learning for Autism identification in Neurological data
Research in Artificial Intelligence and Machine Learning focusing on Neuroscience. Cloud configuration
for parallel computing with GPUs. Usage of modern deep learning frameworks, such as Tensorflow and
Theano.
Advisors: Felipe Rech Meneguzzi PhD, Alexandre Rosa Franco PhD
Evaluation committee: R. Cameron Craddock PhD, Paulo Martins Engel PhD, Rodrigo Barros PhD
GPA: 3.2/4.0
BSc in Computer Science
Universidade de Passo Fundo / Brazil – 2008-2013
Thesis: Artificial Intelligence and the Stock Market: Using neural networks to predict trends
Use of a mathematical model inspired by biological brain functioning to recognise patterns within the
time series obtained from the evolution of a certain stock in BMF & Bovespa.
Advisor: Roberto dos Santos Rabello PhD
Publications
HEINSFELD, Anibal Solon; FRANCO, Alexandre Rosa; CRADDOCK, R. Cameron; BUCHWEITZ,
Augusto; and MENEGUZZI, Felipe. Identification of autism spectrum disorder using deep learning and
the ABIDE dataset. Neuroimage: Clinical, 2017. DOI: 10.1016/j.nicl.2017.08.017
XAVIER, Laura de Lima; HEINSFELD, Anibal Solon; AGUZZOLI, Cristiano; SOLDATELLI, Matheus;
FRANCO, Alexandre Rosa; MENEGUZZI, Felipe Rech; and BUCHWEITZ, Augusto. Classifying brain
states during cognitive tasks: a functional magnetic resonance study in children at reading acquisition
stage. xWorld Congress on Brain, Behavior and Emotions, Buenos Aires, Argentina, 2016.
https://goo.gl/KWmYda
CRADDOCK, R. Cameron; [+18 authors]; HEINSFELD, Anibal Solon; [+26 authors]. Brainhack: a
collaborative workshop for the open neuroscience community, in GigaScience, Vol. 5, 2016. DOI:
10.1186/s13742-016-0121-x. Authors omitted for brevity.
PEREIRA, Ramon F.; MAGNAGUAGNO, Maurício C.; HEINSFELD, Anibal Solon; and MENEGUZZI,
Felipe. LOCUS: An environment description language for JASON, 9th Software Agents, Environments
and Applications School (WESAAC), Niteroi, Brazil, 2015. https://goo.gl/mz9Yd4
Scholarships, Fellowships and Awards
Full Master's degree - Administered scholarship
Pontifícia Universidade Catolica do Rio Grande do Sul / Brazil – 2014-2016.
Funding Institution: CAPES/PROSUP
Advisor: Felipe Rech Meneguzzi PhD
Scientific Initiation Scholarship PIBIC/CNPq
Universidade de Passo Fundo / Brazil – 2010
Funding Institution: CNPq
Project: The hypermedia in the creation of learning collectivities in the municipal schools of Passo Fundo
Advisor: Adriano Canabarro Teixeira PhD
Scientific Initiation Scholarship PIBIC/UPF
Universidade de Passo Fundo / Brazil – 2009
Funding Institution: UPF
Project: The hypermedia in the creation of learning collectivities in the municipal schools of Passo Fundo
Advisor: Adriano Canabarro Teixeira PhD
Predictive Analytics in Mental Health Competition
http://psymri.org/PAC.html – 2016
In this competition, my team and I develop a model to classify patients suffering from Depression and
healthy individuals based on resting-state fMRI data. PAC 2016 provides a balanced dataset of 436
resting-state fMRIs, obtained at 11 different scanning site. The rows of the dataset have been shuffled in
order to anonymize it. Our solution was based on a previous analysis of voxels, in an attempt to identify
the relevant ones. After selecting 200 voxels, we created a correlation matrix and used it as features for
a linear SVM. By using this pipeline, we achieved the second place in the competition.
Multi-Agent Programming Contest
https://multiagentcontest.org/2016/ – 2016
In this competition, my team and I develop a multi-agent system to solve a scenario based on a real city
map, where the goal is to earn as much money as possible, which is rewarded for completing certain jobs.
Our solution is based on JASON to define the behavior of each type of agent and how they coordinate to
form sub teams for tasks, using the Contract Net Protocol. We achieved the first place in the
competition. Full description: https://goo.gl/s94iXJ
Related Additional Coursework
Neural Networks for Machine Learning
University of Toronto/Coursera – 2017
Significant exploration in the field of Artificial Neural Networks. Introduction to basic neural networks
and optimisation methods. Handling of textual, imaging and sequential data. Stochastic, recurrent, deep,
and hierarchical neural nets. Programming tasks using octave.
Computational Neuroscience
University of Washington/Coursera – 2017
Exploration of computational principles of brain functioning, such as vision, sensory-motor control,
learning, and memory. Specific topics covered include representation of information by spiking neurons,
processing of information in neural networks, and algorithms for adaptation and learning. Programming
tasks using Python.
School of Higher Studies - Biomarkers in Neurodegenerative Diseases
Pontifícia Universidade Catolica do Rio Grande do Sul – 2015
Exposition of investigations on new biomarkers of neurodegenerative diseases. The event was attended
by four prominent researchers from the area: Howard Federoff PhD (UC Irvine), Massimo Fiandaca PhD
(UC Irvine), Mark Mapstone PhD (UC Irvine) and Mike Nalls PhD (NIH).
Functional Neuroimaging: A Hands-on approach
Pontifícia Universidade Catolica do Rio Grande do Sul – 2014
Introduction to clinical applications and research for neuroimaging techniques, more specifically
functional magnetic resonance imaging. Fundamentals of neuroimaging techniques, basic equipment
operation and data acquisition, development of studies and research questions and final analysis of
images. Lessons were given by the researchers R. Cameron Craddock PhD, John Wesley VanMeter PhD
and Alessandro Andre Mazzola MSc, as well as the PUCRS researchers Alexandre Rosa Franco PhD and
Augusto Buchweitz PhD.
Machine Learning
Stanford University/Coursera – 2014
Introduction to the field of Machine Learning, Data Mining and statistical recognition of patterns.
Supervised Learning with parametric and nonparametric models, Support Vector Machines, kernels and
neural networks. Unsupervised Learning, through clustering, dimensionality reduction and Deep
Learning. Good practices in Machine Learning.
Research / Teaching Experience
Teaching Assistant
Pontifícia Universidade Catolica do Rio Grande do Sul – 2015
Assist in practical classes to undergraduate students on Artificial Intelligence. Teach classes to students
on Artificial Neural Networks. Develop hands-on activities for students covering topics such as
Automatic Planning, Multi-agent Systems, and Reinforcement Learning.
Neuroimaging of Human Cognition
Pontifícia Universidade Catolica do Rio Grande do Sul – 2015
Development of new techniques for neuroimaging analysis and studies of the neural bases of higher
cognitive processes. Part of the ACERTA project (Evaluation of Children at Risk of Learning Disorder).
Contribution in the construction of predictive models for the diagnosis of disorders, such as dyslexia and
dyscalculia.
Autonomous Systems Group
Pontifícia Universidade Catolica do Rio Grande do Sul – 2015
Research on commitment alignment of multi-agent systems. Assistance for undergraduates on their
research on robotics using Gazebo and ROS. General laboratory management.
Professional Experience
Child Mind Institute
Assistant Research Scientist – 02/2018-Present
Focus on the maintenance of the Configurable Pipeline for the Analysis of Connectomes (C-PAC) project,
by helping other researchers to augment their capabilities on preprocessing and analysis of resting-state
fMRI data.
Planejei
CTO, Chief Data Scientist – 10/2015-Present
Research and management of technologies used in the company. Creation of Marvin, the robot-advisor
of Planejei. Focus on Machine Learning and recommendation systems. Large use of containers, using
Docker and Kubernetes on Google Cloud Platform and Amazon Web Services. Focus on scalability and
continuous integration. Optimization of learning algorithms and messaging methods.
Masters Interval
08/2014-10/2015
Intuitivus Data Processing
CTO – 05/2012-08/2014
Research and management of technologies used in the company. Analyse, develop and deploy
high-performance Business Intelligence (BI) systems to facilitate critical decision making for managers of
customer companies, such as Everything But Water. Focus on analytical information visualisation and
optimization.
Calc Management System
Java/Javascript Developer – 04/2012-08/2013
Development of a taxes calculator for high-trading volume investors. Test-driven programming of
scalable web-portal for thousands of customers, computing millions of stock transactions per day.
Management of load balancing, scaling and general dev-ops of the startup. Integration with Brazilian
investments system, in order to retrieve user transactions automatically.
Compasso Technology
Java/Javascript Developer – 09/2011-04/2012
Development of a large web-portal in Vivo, the largest telecommunications company in Brazil. By using
Oracle technology stack, my team and I developed a support centre, where the user could retrieve its
information about mobile plans, support tickets and online chat.
University of Passo Fundo
PHP/Javascript/Delphi Developer – 02/2011-08/2011
Development of internal systems to the university processes and students portal. Individually, I worked
on an employee scale management system through work orders. By creating an innovative interface for
managers to control the scales, we were able to optimize the maintenance processes of the university.
Other tasks include the proposal of development tools to enrich the systems quality, integration of
payment web-services, bugs fixing, and others.
Volunteer Activities
OpenNeuro: Collaborations on the open-source project, by implementing oAuth2 authentication flow
using ORCID, as an alternative to Google because it is widely used by academics and is not blocked in
China.
FMRIPrep: Collaborations on the open-source project, by implementing parallel workflow execution on
resting-state functional magnetic resonance preprocessing.
Descartaê: Collaborations on the open-source project, by developing a mobile app and a web portal to
employ mechanisms for indexation of waste collection centres.