Tutorial Overview¶
Application Scenarios¶
In the wind power generation business, the weather conditions have a great influence on the power generation, and the business has a strong demand for power generation forecast. This tutorial uses wind power algorithms and sample weather data to simulate actual scenarios, develop machine learning models, and predict the power generation of wind farms.
This tutorial introduces how to use the MI Pipelines to develop, train and deploy machine learning models. The detailed process is shown in the figure below:
The main steps described in this tutorial include:
- Prepare wind power algorithm and external weather data
- Design pipelines by using operators provided by the MI Pipelines
- Build a data arrival monitoring and event triggering mechanism
- Run and publish the pipelines
- Monitor the model indicators
Prerequisites¶
- The wind power algorithm file (speed_to_power.csv) and external weather data accumulated in the field of wind power generation have been obtained.
- You have understood the functions and usage methods of various operators in EAP MI Pipelines. For more information, see Operator Reference.
- You should ensure that the organization has requested Batch Processing - Queue, Data Warehouse Storage, and File Storage HDFS resources through the Resource Management page for storing and processing data required for model training.
- You should ensure that the organization has request the ML Model - Container resource through the Resource Management page. For more information, see Resource Management on EnOS.
- You should have completed the connection configuration of the Git data source and Hive database through the EAP Resource Configuration > Connection Configuration page.