Answering 59 scikit-learn questions (video)
This article is originally published at https://www.dataschool.io/

In April, I hosted a live, public webcast to answer any questions about scikit-learn. I've been teaching scikit-learn for five years, and I really enjoy sharing my knowledge with others!
Hundreds of people attended, and I answered 59 questions in 90 minutes! ⏱????
You can watch the recording right now:
???? Watch me answer 59 scikit-learn questions LIVE ????
Here are a few of the questions I answered:
- How do I include a categorical feature in a model?
- How should I deal with class imbalance?
- Can I do all of my preprocessing in scikit-learn (instead of pandas)?
- When should I standardize my features?
- Should I split my dataset into train/test OR train/test/holdout?
- Is it worth my time to learn scikit-learn, since deep learning has been so successful?
- How do I know if I have "enough" data to build a model?
- What is data leakage?
Click on a question to jump directly to my answer!
New scikit-learn course ????
I also answered questions about my new course, Building an Effective Machine Learning Workflow with scikit-learn. Here are a few of those questions:
- Would you recommend the course to someone who is new to ML?
- How different is the course from your YouTube videos?
- Can you explain the "Live Course + Advanced Course" bundle?
- How can I best prepare for your course?
Although the "Live Course" has already taken place, you can still enroll today to access 8 hours of recordings ???? and my detailed instructional notebooks ????
???? Learn more and enroll here ????
"This was one of the best data science classes I have ever taken... I was impressed with Kevin's easy-to-understand teaching style where he clearly explains the 'what' and 'why' of each principle... I highly recommend this course." - Khaled Jafar (Director of Analytics)
Please comment below or email me if you have any questions!
Thanks for visiting r-craft.org
This article is originally published at https://www.dataschool.io/
Please visit source website for post related comments.