Train Custom Machine Learning Models with No Data Science Expertise
summary
Cloud AutoML on Google Cloud lets developers upload labeled images, tables, or text and get a hosted model and REST endpoint without writing TensorFlow. Demos use Cloud Vision, Video Intelligence (auto transcription and bounding boxes), Natural Language, Speech and Translation; a clothing object‑detection model is built from a Kaggle dataset plus additional photos annotated via a data labeling service with PDF instructions. Training shows “check back in four hours,” Tables matches median competition performance in an hour and improves with 12–24 hours; models work with AutoML Tables and even with under 200 examples.
AI-written summary · about
chapters (13)
- 0:00Intro
- 0:54Why Machine Learning
- 1:47What is Machine Learning
- 7:07Video Intelligence API
- 11:02Label Data
- 12:27Building a Model
- 15:43Neural Networks
- 17:21Data Types
- 18:06Data Science Competition
- 18:46AutoML Results
- 19:20Custom Natural Language
- 20:12Custom Entity Extraction
- 20:50Summary
related talks
42mUse Machine Learning in your code, without being a ML expert
57mBuilding smarter solutions with Machine Learning, from magic to reality
47mBuilding smarter solutions without machine learning expertise
42mMachine Learning APIs, from Training [...]
44mBuilding smarter solutions with no expertise in machine learning
47mBuilding smarter apps with Machine Learning, from magic to reality
36mCloud AI Services: What they are and how to use them
54mBuilding smarter apps with Machine Learning, from magic to reality