Extremely excited in achieving ML Engineering certification as I embark on my AI/ML journey with Google Cloud. There is not a lot of material available while preparing for this certification as the exam is released less than a month ago.
Study guide has most of the content on what to focus on. This exam is for ML Engineers. I would suggest anyone planning to take the exam first pursue the Data Engineering Exam(Prep content here) as the focus is more on data.
Key topics to focus on#
- Orchestration of the Machine Learning life cycle
- Data Pre-processing, preparation and options
- Feature Engineering strategies
- Training and Deploying techniques , options - Pros and Cons
- ML Ops and its workflow
- Spectrum of options available - API, Auto ML , DIY and AI Platform
- Understanding Gradient Descent, Loss , Regularization
- High level understanding of Regression, Classification, Clustering, CNN, DNN, RNN
I used the below preparation contents. Hope it is of some help to you. Good luck with your exam. And thank you to the folks who helped me in identifying the key areas to focus on.
Machine Learning Crash Course - Google - MUST read even if you are not taking the exam
Have a good understanding of Google Cloud Services - Data prep, Data fusion, Data flow, Composer, API, Auto ML, AI Platform (all features), Bigquery ML
Google Cloud Solutions - Wealth of content with architectures#
Data Pre-processing#
Analyzing and validating data at scale for machine learning with TensorFlow Data Validation
Building production-ready data pipelines using Dataflow: Deploying data pipelines
Data preprocessing for machine learning: options and recommendations
Data preprocessing for machine learning using TensorFlow Transform
Considerations for sensitive data within machine learning datasets
Machine learning with structured data: Data analysis and prep
Training and Prediction#
Comparing ML models for Predictions using Dataflow pipelines
Best practices for performance and cost optimization for Machine Learning
Minimizing real time prediction serving latency in Machine Learning
Optimizing Tensorflow models for serving] (@Lukman Ramsey)
MLOps#
MLOps: Continuous delivery and automation pipelines in machine learning
Setting up MLOps Environment on Google Cloud
Architecture for MLOps using TFX, Kubeflow Pipelines, and Cloud Build
Coursera content - Good Refresher#
(Valentine Fontama, Valliappa Lakshmanan)#
Other blog posts with relevant content#
Special Thanks to Steve Walker, Sanjay Agravat, Fernando Sanchez, Amit Rai, Michael Ross, Jamin Solensky and Yogesh Tiwari