-
Story
-
Welcome
- Set up
-
Access to the CDF project
-
What this notebook will achieve
-
Google Colab Set Up
-
Getting started
-
Let's walk through the notebook
- Soft sensor
-
Intro
-
Notebook explained, part 1
-
Notebook explained, part 2
-
Notebook explained, part 3
-
Notebook explained, part 4
- Anomaly detection
-
Intro
-
Notebook explained
-
End of course
-
Feedback
Data Modelling
In this course you’ll continue in the role of an engineer in charge of operating a gas compressor, and will learn how to deal with live data in data modelling.
Welcome to the course on data modelling
In this course, you’ll continue in the role of an engineer in charge of operating a gas compressor. Unfortunately, the output pressure sensor has broken. To predict the missing sensor data, you'll go through several steps in order to create and compare various soft sensor models that are able to predict what this sensor would be in the future. Then, you’ll build a simple anomaly detection model based on this work, in order to predict when the compressor isn’t operating as normal.
You will learn how to:
- Extract live data from an oil rig in the North Sea.
- Visualize and inspect data directly from Cognite Data Fusion.
- Apply and assess several machine learning models for time series prediction.
Prerequisites
Access to Open Industrial Data and Google Colab. If you would prefer to run code cells outside of Google Colab, you can download the notebook or copy the code directly from the course. This course is part of the learning path "Getting hands-on with data science".
Instructors
This course was developed by Cognite Academy.