- Level Expert
- Duration 12 hours
-
Offered by
About
In the first course of Machine Learning Engineering for Production Specialization, you will identify the various components and design an ML production system end-to-end: project scoping, data needs, modeling strategies, and deployment constraints and requirements; and learn how to establish a model baseline, address concept drift, and prototype the process for developing, deploying, and continuously improving a productionized ML application. Understanding machine learning and deep learning concepts is essential, but if you're looking to build an effective AI career, you need production engineering capabilities as well. Machine learning engineering for production combines the foundational concepts of machine learning with the functional expertise of modern software development and engineering roles to help you develop production-ready skills. Week 1: Overview of the ML Lifecycle and Deployment Week 2: Selecting and Training a Model Week 3: Data Definition and BaselineModules
The Machine Learning Project Lifecycle
1
External Tool
- Intake Survey
4
Videos
- Welcome
- Steps of an ML Project
- Case study: speech recognition
- Course outline
1
Readings
- [IMPORTANT] Have questions, issues or ideas? Join our Forum!
Graded Assessment
1
Assignment
- The Machine Learning Project Lifecycle
Deployment
4
Videos
- Key challenges
- Deployment patterns
- Monitoring
- Pipeline monitoring
1
Readings
- Week 1 Optional References
Graded assessment
1
Assignment
- Deployment
Lecture Notes (Optional)
1
Readings
- Lecture Notes Week 1
Ungraded Lab
2
Labs
- Deploying a Deep Learning model
- Deploying a deep learning model with Docker and a cloud service (optional)
Select and train a model
1
Assignment
- Selecting and Training a Model
5
Videos
- Modeling overview
- Key challenges
- Why low average error isn't good enough
- Establish a baseline
- Tips for getting started
Error analysis and performance auditing
4
Videos
- Error analysis example
- Prioritizing what to work on
- Skewed datasets
- Performance auditing
Data iteration
7
Videos
- Data-centric AI development
- A useful picture of data augmentation
- Data augmentation
- Can adding data hurt?
- Adding features
- Experiment tracking
- From big data to good data
1
Readings
- Week 2 Optional References
Graded assessment
1
Assignment
- Modeling challenges
Lecture Notes (Optional)
1
Readings
- Lecture Notes Week 2
Ungraded Lab
1
Labs
- A journey through Data
Define Data and Establish Baseline
7
Videos
- Why is data definition hard?
- More label ambiguity examples
- Major types of data problems
- Small data and label consistency
- Improving label consistency
- Human level performance (HLP)
- Raising HLP
Label and Organize Data
1
Assignment
- Data Stage of the ML Production Lifecycle
4
Videos
- Obtaining data
- Data pipelines
- Meta-data, data provenance and lineage
- Balanced train/dev/test splits
1
Readings
- [IMPORTANT] Reminder about end of access to Lab Notebooks
Scoping (optional)
1
Assignment
- Scoping (optional)
5
Videos
- What is scoping?
- Scoping process
- Diligence on feasibility and value
- Diligence on value
- Milestones and resourcing
1
Readings
- Week 3 Optional References
Lecture Notes (Optional)
1
Readings
- Lecture Notes Week 3
Ungraded Lab
1
Labs
- Data Labeling
Final Project
1
Labs
- The Machine Learning Project Lifecycle
1
Videos
- Final project overview
Course Resources
1
Readings
- References
Acknowledgements
1
Readings
- Acknowledgments
Auto Summary
Dive into "Introduction to Machine Learning in Production," an expert-level course on Coursera focused on Data Science & AI. Guided by industry experts, you'll learn to design and deploy ML systems end-to-end, from project scoping to addressing concept drift. Over 720 minutes, this course covers the ML lifecycle, model training, and data baselines. Ideal for those seeking to enhance their AI career with production engineering skills. Subscription options include Starter, Professional, and Paid plans.