ML for the GCP PDE Exam

#Google Cloud #Professional Data Engineer Certfication #Machine Learning

Intro

This post is part of a series of posts with notes as I’m studying for Google’s Professional Data Engineer Certification.

This particular post covers ML topics.

Disclaimer

Please read this disclaimer.

Key Concepts

  • L1 vs. L2 regression:
    • L1 estimates the median of the data using absolute value
      • Reduces low-value features
      • Robust to outliers
      • Good when only certain features contribute to success of model
    • L2 estimates the mean of the data to avoid overfitting
      • Not recommended for feature selection
      • Good when all features contribute relatively equally to the success of a model
  • TensorFlow
    • Know how TensorFlow can be deployed (also, cost vs. value)
      • CPUs, GPUs, TPUs
  • BQ ML
    • Understand the basic flow when using BQ ML:
  • Vertex AI
    • Create computation graph and training app
    • Package app
    • Start Vertex AI job to run packaged app