In a world where the amount of available data is increasing at lightning speed, using machine learning and data science is the most natural way for improving most analytical tasks and moving to much more flexible solutions compared to traditional heuristics and business rules.
When tasked with developing a predictive model, your objective is most likely to help solve a specific business problem. This can require not only having to cooperate with developers and other data scientists on a pathway for delivering the technical solution itself, but also convincing your boss and stakeholders of the necessity for using applied data science methods. At this point, not only do you need to master applied methods and algorithms, but also understand how to align them with real business needs.
In this webinar, Vladimir Mikhnovich will help you overcome this by discussing:
- Key differences between communicating results to fellow data scientists, analysts and stakeholders
- Why you shouldn’t aim for building too complex models in favor of simple yet understandable ones
- How to use proper business metrics and cost-sensitive approach for evaluating models performance and fine-tuning them