第 11 屆 iThome 鐵人賽

AI & Data

Predicting Inter Bus Arrival Times 系列 第 8

Day 8 我要成為 Googler +2


  1. How Google does Machine Learning
  2. Launching into Machine Laerning

先 How Google does Machine Learning 後 Launching into Machine Laerning

1 min Diane Greene,CEO,Google Cloud & Lak

enable Python programmers to do machine learning and data scientists to build prroduction-ready machine learning models
help reduce organizational silos. because of the combination of global serverless infrastructure like BigQuery and open source software like TensorFlow

目的: give us a practical real-world introduction to machine learning
產品行銷 : BigQuery

5min Specialization Agenda
what Google means when we(Google) say we(Google) are AI first ?

big picture is very important

  1. creating a good Machine Learning dataset
  2. building your first Machine Learning model with Tensorflow

how to train,deploy,and predict with ML models in a way that they're
production ready --> delve back into Machine Learning theory -->
powerful recommendation system

由 Google engineers teach how approach new and existing products,how Google design products

ml products 的成長曲線 ?
2012 : 0 model
2016 : 4000 models

how our engineers infuse ml into products

推薦平台 : Google Cloud Platform
pros : scalability,reliability

right handside : cost lots of time
left handsise : ml serving

how to build batch and streaming data pipelines ?


  1. related to recommendation system
  2. 我忘記啦

    當時 應該截圖建檔的,現在記憶有限...


圖片來源 :

  • GCP Team

Day 7 我要成為 Googler +1
Day 9 我要成為 Googler +3
Predicting Inter Bus Arrival Times 30