A practical training Clean Machine Learning for programmers, statisticians, DevOps, researchers, and other similar practitioners.
The following topics are covered:
- The pipe and filter design pattern: an architectural style.
- Rethinking how to program machine learning applications: using pipelines.
- Two practical exercices / coding kata explained in depth: Machine Learning Pipelines and their abstractions - with solution. Learn how to structure your machine learning code into a neat pipeline by practice. Here is the code for the Clean Machine Learning kata, as a preview.
- Automated Machine Learning: learn how to automate the training loop and how to tune your hyperparameters.
- Software architecture: learn how to structure your machine learning projects.
The concepts explored in this training in Python are reusable to make things work as flexibly as possible with the most popular machine learning libraries such as scikit-learn, TensorFlow, PyTorch, Keras, and Neuraxle. This way, it'll be possible to merge those technologies into your machine learning pipeline or app. It is known that merging those technologies is difficult, and our well-recognized way of structuring code into a pipeline will allow you to not hit a wall when it comes to build applications that will be deployable for production.
Note: this 3-hours online video training will be delivered within 7 days from the purchase due to processing delays.
Contact us for in-person delivery for your group: available in Québec only, and add $1000 in fees, also available in French, we must also be available at the suggested time 2 weeks in advance.
A LinkedIn certificate is available upon passing the course.