Heitor Gomes
Research Scientist
Wellington, Wellington, New Zealand
Summary
👤
Senior🎓
Top SchoolHeitor Gomes is a research scientist and seasoned machine learning engineer with a decade of experience specializing in streaming data and online learning. Currently a postdoc and adjunct consultant in Wellington, he develops research and open-source tooling—most notably contributions to MOA and StreamDM—helping bridge academic advances and production-grade stream mining on platforms like Apache Spark. His work spans algorithmic improvements (drift detection, evaluators) and practical engineering fixes that improve runtime estimation and multi-class metrics for real-time systems. Heitor combines rigorous research with hands-on back-end implementation, and quietly excels at tightening core library components that make streaming ML robust and reproducible.
10 years of coding experience
Pontifical Catholic University of Paraná
English, Portuguese, Spanish, French
Github Skills (15)
data-mining10
javas10
machine-learning10
machine-learning-algorithms10
spark-streaming10
clustering10
evaluation10
classification10
metric10
java10
scala9
algorithms8
data-structures8
algorithm8
data-structure8
Programming languages (4)
Github contributions (5)
huawei-noah/streamDM
Aug 2017 - Jul 2019
Stream Data Mining Library for Spark Streaming
Role in this project:
Data Scientist
Data ScientistContributions:47 commits, 33 PRs, 26 pushes in 1 year 11 months
Contributions summary:Heitor focused on enhancing the `BasicClassificationEvaluator` within the `streamdm` library, adding runtime estimation, and multi-class evaluation metrics, improving its overall functionality. They implemented calculations for metrics such as precision, recall, Fbeta-score, and specificity, enriching the evaluation capabilities of the system. Furthermore, the user addressed parameter settings and made formatting changes within the evaluation code. These changes likely aimed to provide more comprehensive and informative results for stream data mining tasks.
miningdata-streamdata-miningstreamsstreaming
Waikato/moa
Apr 2016 - Apr 2022
MOA is an open source framework for Big Data stream mining. It includes a collection of machine learning algorithms (classification, regression, clustering, outlier detection, concept drift detection and recommender systems) and tools for evaluation.
Role in this project:Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:2 reviews, 24 commits, 36 PRs in 6 years
Contributions summary:Heitor primarily contributed to the MOA framework, a library for data stream mining, by modifying and enhancing core functionalities. They focused on refining existing classes like `Instances` and `OzaBagASHT`, by addressing constructor issues and enforcing the use of `ASHoeffdingTree` as the base learner. Their work also included adding and modifying classes relevant to machine learning techniques, particularly drift detection. Further contributions involved adding new evaluators, such as DelayedLabelingEvaluators.
pythondata-streamstreamdriftclassification
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.