ROBSON CORDEIRO
Data Scientist

robsonc@andrew.cmu.edu
510-519-0845
linkedin.com/in/rlfcordeiro
https://github.com/robsoncordeiro/robson-portfolio
Pittsburgh, PA

Summary
======
High-performing Data Scientist with a Ph.D. degree in Computer Science and experience in both industry and academia. Expert in Database SQL, fraud detection, outlier detection, classification, and clustering.


PROFESSIONAL EXPERIENCE
======
Innovation Consultant
PNC Bank
2024 - Ongoing
Pittsburgh, PA
 - Created a fraud detection tool that can increase income considerably
 - Identified features that largely improved the percentage of fraud detected by a model

Project Scientist
Carnegie Mellon University
2024 - Ongoing
Pittsburgh, PA
 - Found 2 unusual behaviors in financial data that were previously unknown to domain experts, catching hundreds of fraudulent transactions that would go otherwise unnoticed
 - Created a tool to visually inspect unusual behaviors in financial data, which allowed domain experts to deem as fraudulent or suspicious 2 behaviors identified previously

Data Science Researcher and Professor
Universidade de Sao Paulo - USP
2013 - 2023
Brazil
 - USP is the premier university in Latin America
 - Created over 20 innovative machine learning tools. It includes software for multimodal data classification, visualization, fraud and anomaly detection, and clustering applied to finance, ecommerce, cybersecurity, and social networks
 - They all outperform the best competitors in efficiency or effectiveness by being up to 4 orders of magnitude faster or up to 50% moreaccurate, as demonstrated through extensive experimental evaluation and verified by independent peer reviewers
 - Taught 33 courses on SQL-based modeling, querying, and manipulation of databases and data warehouses, software engineering, machine learning, algorithms, data structures, and object-oriented programming to graduate and undergraduate students
 - Mentored 11 Ph.D. or M.Sc. graduate students with 100% success rate


ACHIEVEMENTS
======
Created 20+ Machine Learning Tools
It includes fraud detectors for financial data, and surveillance software used in practice for U.S. national security

Published 40+ articles in top venues
40+ articles published in world-class data science int. conferences and journals, like ACM's KDD and CIKM, IEEE's ICDM, ICDE and TKDE, SIAM SDM, and ECML / PKDD

Best CS Dissertation Award
Won first place out of all Computer Science Ph.D. Works concluded over the year in Brazil, as evaluated by the Brazilian Computer Society


TECHNICAL SKILLS
======
General
 - Machine Learning 
 - Artificial Intelligence
 - Data Mining
 - Software Engineering
 - Databases
 - Cloud Computing

Languages/Tools
 - Python
 - Java
 - C++
 - SQL
 - Cypher
 - Git

Machine Learning, Stats and Experimentation
 - Classification (XGBoost
 - Random Forest
 - AutoGluon)
 - Clustering (K-means
 - DBSCAN
 - EM)
 - Anomaly detection (iForest
 - KNNOutlier
 - LOF)
 - Principal Component Analysis
 - Time series forecasting (ARIMA)
 - Embedding (t-SNE
 - UMAP)
 - Visualization (Charts
 - Histograms
 - Heat maps
 - Box plots
 - Word clouds
 - Critical difference diagrams)
 - Hypothesis testing
 - Covariance and correlation modeling

Libraries/Frameworks
 - scikit-learn
 - MLlib
 - SciPy
 - Spark
 - Pandas
 - Oracle
 - Neo4j
 - NetworkX
 - Cytoscape.js
 - Streamlit
 - Matplotlib
 - seaborn
 - NumPy
 - MATLAB


Education
======
Post-Doctorate, Computer Science
Universidade de Sao Paulo - USP
Brazil

Ph.D., Computer Science
Universidade de Sao Paulo - USP
Brazil

M.Sc., Computer Science
Universidade Federal do Rio Grande do Sul - UFRGS
Brazil

Bachelor of Science, Computer Science
Universidade do Oeste Paulista - UNOESTE
Brazil