Data Science: type safety & A/B testing
New year, a new season of tech talks! Save your spot for another session from our online series. This time we invited experts in #Data Engineering, #Data Science & #Machine Learning from Product Madness and VirtusLab to share their expertise with us.
The event will take place on the 27th of January, 2022.
We start at 6:00 PM.
Don’t miss your chance and save a spot!
Agenda
pandas-stubs — How we enhanced pandas with type annotations
The “type stubs” mechanism allows you to perform static type analysis for non-annotated Python code. We will present the “pandas_stubs” library, which enriches the pandas code with the missing type information. We will describe its genesis and development process as well as the current state of the project.
We will show examples where its use allows us to avoid hidden errors in the code. We will talk about the collection of the library, current use, and development plans.
A/B Testing Madness
Many companies today use A/B testing to support and improve their decisions. A/B tests are studies when a target audience is split into several groups, each seeing a specific version of a product. The goal of the A/B tests is simple – to decide what works better. There are many challenges with A/B tests – mathematical, technological, and operational.
Which statistical procedures to use for inferences? How to incorporate A/B testing into technological procedures? How should businesses act on the results of the analyses? The talk will cover how the Data Science team addresses these issues at Product Madness – a mobile gaming company. The mobile gaming industry has plenty of data, a dynamic user base, and a fast-paced business, which makes A/B testing exciting and challenging
