Practical introduction to machine learning
What you'll learn, and how you can apply it
- Discover what it means to "use ML" and how to go from where you are to following through on a project
- Understand the intention and mental model that achieved results with new algorithms
Who is this presentation for?
- You're a CTO, software engineer, or business analyst.
- Experience with Python or programming
Hardware and/or installation requirements:
- A laptop with access to JupyterHub or Python 3, GitHub, Keras, CatBoost, and scikit-learn installed
Introduction and warm-up
- Discuss how motivated you are to learn ML and why you haven’t integrated ML into your daily work
- Hear knowledge gained from 12 years of real-world experience
- Understand what ML is without deep math, but with in-depth thinking and work
What is ML?
- Explore the nuances of how ML algorithms turn data into insight and the distinction between AI and deep learning
- Learn the classes of ML and how to approach building a project
- Hands-on exercise: Wordball
Supervised learning: Regression and classification
- Delve into two examples of supervised learning, predicting house values and classifying spam
- Explore the TACT method to ML project mastery, how to compose a model via training and adequate tests, how to interpret and code up good examples, and how to pick useful algorithms
Taking it to 11: Feature engineering and ensemble learning
- Learn techniques to improve models and revisit concepts, namely around the curse of dimensionality
- Attempt to classify handwritten numbers using the techniques you have already, and find that it quickly fails
- Dig into feature engineering (matrix factorization and unsupervised methods) and ensemble learning, including state-of-the-art algorithms like XGBoost and CatBoost and explainable boosting machines
- Explore the use and abuse of matrix factorization, how to cluster data together, what unsupervised learning is attempting to do and why it matters, “features,” why ensemble learning works so well, where ensemble learning gets used extensively, what random forests are, and what bagging and boosting are
Deep learning dip
- Dip your toes into neural nets and deep learning—you’ll learn enough to be dangerous
- Determine if you can do any better using your MNIST dataset, including feed forward neural networks, activations, recurrence, convolutions, and why you should avoid using deep learning most of the time
Wrap-up and Q&A
About your instructor
Matt Kirk is the founder of Your Chief Scientist, a firm devoted to training small cohorts of highly motivated engineers to become data scientist practitioners. He pulls from his experience writing Thoughtful Machine Learning with Python as well as his clients like ClickFunnels, Garver, SheerID, SupaDupa, and Madrona Venture. To find out more, check out yourchiefscientist.com.
Get a Training Pass to add this course to your package.
Leave a Comment or Question
Help us make this conference the best it can be for you. Have questions you'd like this speaker to address? Suggestions for issues that deserve extra attention? Feedback that you'd like to share with the speaker and other attendees?
Join the conversation here (requires login)
For conference registration information and customer service
For more information on community discounts and trade opportunities with O’Reilly conferences
For information on exhibiting or sponsoring a conference
For media/analyst press inquires