Advanced Model Optimization
About the course
Optimizing your models allows you to boost their impact by ensuring they are robust and deliver the highest possible performance. Follow our innovative two-day Advanced Model Optimization course and have our trainers walk you through the key steps and real-life exercises. We will begin with mastering approaches to tune a model’s hyperparameters, such as Bayesian Optimization. Followed by automating the feature selection process and increasing a model’s robustness, using techniques like Genetic Algorithms, cross-validation, and quantile regression. Finally, we will focus on identifying and dealing with data shifts to safeguard your model in production. Significant time is spent on covering multiple advanced methods in the model optimization workflow, highlighting often incorrectly applied methods and mistakes ensuring you are prepared for anything. After discussing the theory, each method comes with a practical case to get your (Python programming) hands dirty and implement the method in a case setting. We end the last day with a modeling battle, where all participants can earn points for their model optimization efforts in a real-life regression case. The winner can call himself a true modeling hero. Will you be that hero?
Why this is for you
When creating a model, you want to optimize it in order to realize maximum performance. You’ve probably done so yourself, but didn’t you find it time-consuming and requiring much manual tweaking? Indeed, model optimization can be a cumbersome task. However, it doesn’t need to be.
In this training, we teach you how to work smarter, not harder. You will get firsthand experience with the best-practices of efficient and (partly) automated model optimization. Moreover, we pay special attention to model robustness, such that your models are future-proof and have long-term business value. After this course, your models will perform better with less developing efforts. What’s more to like?
This is a professional level training for Data Scientists who have completed badges Machine Learning Process (3201), Classification Using Tree Models (3202), and Regressions Models (3203). It involves tough coding challenges and various abstract topics which means you need to have a good understanding of the Machine Learning modeling process and how to create a model in Python.
What you’ll learn
- The standard and advanced methods to tune a machine learning model’s hyperparameters
- How to implement two core variable selection strategies: Filter and Wrapper
- How to assess and increase model performance
- How to statistically identify data shifts
- Methods to make a model more adaptive and less susceptible to a changing dataset
Theory and practical use
- Hyperparameter tuning – Master hyperparameter tuning to deliver maximum performance
- Feature selection – Automate the feature selection procedure for a faster modeling process and better performance
- Robustness – Evaluate and increase a models performance
- Data shifts – Capable of identifying and dealing with data shifts
All trainings in the GAIn portfolio combine high-quality standardized training material with theory sessions from experts and hands-on experience where you directly apply the material to real-life cases. Each training is developed by top of the field practitioners which means they are full of industry examples along with practical challenges and know-how, fueling the interactive discussions during training. We believe this multi-level approach creates the ideal learning environment for participants to thrive.