Data Science on Google Cloud Platform: Predictive Analytics

40mIntermediate2018-11-07

Authors

Kumaran Ponnambalam

Working with data for 20+ years

Course details

Predictive analytics use historic data to look forward, enabling organizations to make better decisions. However, making accurate predictions from big data can be an overwhelming task. Enter Google Cloud Platform (GCP), a suite of cloud-computing services that bring scalability, elasticity, and automated machine learning to predictive analytics. This course—one of a series by data scientist Kumaran Ponnambalam—shows how to apply the power of GCP to generate predictions for your business. Start off by exploring the different tools and features for predictive analytics in GCP, including Cloud Dataproc, Cloud ML Engine, and the machine learning APIs such as Cloud Translation, Cloud Vision, and Cloud Video Intelligence. Then explore learn how to build, train, and deploy models to create predictions. Plus, learn best practices for cost control, testing, and performance monitoring of predictive models.

Learning objectives
Evaluating the machine learning tools in GCP
Understanding the predictive analytics process
Building models
Training models with jobs
Building and running predictions
Best practices for cost control, testing, and performance monitoring

Skills covered

Data ModelingGoogle CloudSoftware Development ToolsGoogleCloud PlatformsCloud ComputingData ScienceSoftware DevelopmentOne-Off

Concepts

0. Introduction

01 - Why use predictive analytics on GCP
02 - Data science modules covered

1. ML Options in GCP

03 - Cloud Dataproc
04 - Cloud ML Engine
05 - Cloud Natural Language
06 - Cloud Translation
07 - Cloud Vision
08 - Cloud Video Intelligence
09 - Cloud Dialogflow

2. Cloud ML Basics

10 - Models
11 - Model versions
12 - Jobs
13 - Predictive analytics process

3. Model Building with Cloud ML

14 - Understanding input data
15 - Build and test model locally
16 - Upload files to Cloud Storage
17 - Modify code to work with GCP
18 - Creating a training package
19 - Running training synchronously
20 - Training using jobs

4. Predictions in Cloud ML

21 - Creating a deployment model
22 - Creating a model version
23 - Creating a prediction dataset
24 - Running a prediction

5. Cloud ML Best Practices

25 - Cost control
26 - Local testing
27 - Performance monitoring

Conclusion

28 - Next steps