12-05, 19:00–19:30 (UTC), Data/ Data Science Track
This talk showcases and exemplifies the rapid specification and execution of Quantile Regression workflows. Various use cases are discussed, including fitting, outlier detection, conditional CDFs, and simulations, using different types of time series data.
Quantile Regression (QR) is a powerful analysis method that is often considered superior to other regression techniques. The benefits of QR are highlighted in this talk through multiple examples from various "real-life" time series, such as finance, weather, and human activities. The use of QR for fitting, outlier detection, conditional CDFs, and simulations will be demonstrated. The analysis is significantly simplified with the new Python package "Regressionizer," which allows for quick setup and execution of QR workflows.
This presentation is intended for data analysts, data scientists, engineers, and anyone with an interest in time series analysis. A basic understanding of Python is all that is required from the audience, as the coding pipelines have been designed to be straightforward and easy to follow.
No previous knowledge expected
I am an applied mathematician (PhD) with 30+ years of experience in algorithm development, scientific computing, mathematical modeling, natural language processing, combinatorial optimization, research and development programming, machine learning, and data mining.
In the last 16 years, I focused on developing machine learning algorithms and workflows for different industries (entertainment, recruitment, healthcare, manufacturing, logistics.)
I am a former kernel developer of Mathematica.