About Me

Who Am I?

> Hi! I'm Renato, Data Scientist and PhD student in Artificial Intelligence at the Federal University of Rio de Janeiro, Brazil. I currently work as a Data Scientist at Petrobras (short for "Petróleo Brasileiro S.A."), one of the largest Oil & Gas companies in the world, working on deep learning for Oil & Gas Exploration.

As a BSc in Physics and MSc in Applied Mathematics, I have background in Linear Algebra, Statistics, Probability and much of the formal knowledge necessary to understand and apply machine learning methods. Also, through my business and academic track, I was able to develop soft and hard skills for working either in specialized or interdisciplinary teams.

My technical interests cover a wide range of techniques in statistical learning and advanced analytics, with special emphasis on deep learning methods and applications.

Get CV

Some tech I manage to work with

Python

Git

SQL

Julia

R

Heroku

Docker

Flask

HTML

Education

@Federal University of Rio de Janeiro

Ongoing. Expected graduation: 2023.

Working on zero-shot learning and lifelong learning methods.

Relevant coursework:

  • Neural Networks
  • Deep Learning
  • Continual Learning
  • Information Theory

The Deep Learning Specialization from DeepLearning.ai is a five-course specialization that helps understand Deep Learning fundamentals and applications. The specialization was created and is taught by Dr. Andrew Ng, from Stanford University, a global leader in AI and co-founder of Coursera.

Coursework:

  • Neural Networks and Deep Learning
  • Improving Deep Neural Networks
  • Structuring Machine Learning Projects
  • Convolutional Neural Networks
  • Sequence Models
@Fundacao Getulio Vargas

Emphasis on statistical learning. For the thesis, I have studied properties of the macroscopic behavior of bus network traffic in the city of Rio de Janeiro, Brazil.

Relevant coursework:

  • Machine Learning
  • Statistics
  • Probability
  • Linear Algebra
  • Information Retrieval
  • Data Visualization
@Federal University of Rio de Janeiro

Studied magnetic properties of materials at low temperatures, in the area of condensed matter physics.

Some Projects

>> Neural Networks Implementations - Brief examples of a regression, a binary classification and a multi-class classification problem, solved with neural networks. Tools: Python, Keras/TensorFlow.

>> Churn prediction - Binary classification model, applied to customer churn prediction problem. Model deployed in production here. Business problem: Churn occurs when customers stop using a company’s service. So, by predicting churn, companies can develop personalized customer retention campaigns. Tools: Python, ScikitLearn, Pandas, Numpy, Flask, Heroku.

>> Malaria detector - Convolutional neural network classification model, applied to Malaria identification in cell images, using data taken from the official NIH Website. Tools: Python, Pandas, Matplotlib, Seaborn, Keras/TensorFlow.

>> Default prediction - Neural network classification model, applied to a default prediction problem from Lending Club company, using data available from Kaggle. Business problem: when the company receives a loan application, it has to make a decision for loan approval. Tools: Python, Pandas, Seaborn, Keras/TensorFlow.

>> Python Web Scraper - Web Scraper developed for automatic retrieval of CVM data. (CVM is an autonomous entity, linked to Brazil's Ministry of Economy, with the objective of regulating securities market in Brazil). Tools: Python, BeautifulSoup, Selenium.

>> Object recognition in images - Image classification with convolutional neural networks on CIFAR-10, a dataset of 50.000 32x32 color training images, labeled over 10 classes. Tools: Python, Keras/TensorFlow, Matplotlib, ScikitLearn.

>> Breast cancer survival - The dataset describes breast cancer patient data and the outcome is patient survival (Haberman's Survival Data Set). The goal is to develop a classification model to help predicting survival, given a few features. Tools: Python, Keras, ScikitLearn, Seaborn.

>> Spotify Clustering - Small project using Spotify open API to retrieve playlist data and cluster its musics according to loudness and duration. Tools: Python, ScikitLearn, Pandas.