Dr. Pavel Silveira

Machine Learning Reply DE (formerly part of Data Reply GmbH)

- Developing machine learning projects across a variety of industries - Event forecasting, costumer behavior analysis, digital KPI's improvement - Machine Learning roadmap, time series analysis, Deep Learning, LSTM, ML on the cloud - Presenting findings to stakeholders and providing business recommendations - Delivering hands-on and C-level data science trainings for companies

- Advanced eight-week Data Science bootcamp using real-world data to build predictive and classification models - Web scraping (beautiful soup) and data wrangling (pandas, numpy, scipy, matplotlib) - Machine Learning (scikit-learn): linear and logistic regression, nearest neighbors, decision trees, random forrest, svm, k-means, neural networks - NLP, digital signals and time series prediction - Distributed computing: Hadoop and Apache Spark

- Analyzing nonholonomic systems and sub-riemannian models through geometrical and analytical properties - Delivered exercise classes on several Analysis courses and co-supervised master's thesis works

- Conducted research in optimization and control, differential and metric geometry, and dynamical systems - Solved a classical, albeit still unanswered, engineering-inspired problem about optimal solutions of linear quadratic models - Results published in specialized journals, and presented in conferences and workshops in Europe - Webmaster of the Mathematics Area website for two years

- Designed and implemented machine learning algorithms and statistical models for large volume of biological data, e.g. DNA sequences - Focused on Bayesian networks, metaheuristic-stochastic optimization, non-parametric experimental design and signal processing - Conducted exercise groups and delivered lectures on Mathematical Analysis, Geometry and Topology - Acted as coordinator for the Cuban Mathematical Olympiad for University Students and delivered training classes for mathematics competitions