Discovery Guides - Machine Learning and Rules
This discovery guide gives an overview of how you can use our machine learning features to categorise and forecast data or apply business rules for event-driven automation of tasks.
This discovery guide gives an overview of how you can use our machine learning features to categorise and forecast data or apply business rules for event-driven automation of tasks.
Machine Learning Overview
The machine learning module within OpenDataDSL makes it easy for you to create, teach, test and use models to categorise, discover and forecast your own data.
The process consists of the following stages:
- Prepare your 'marked-up' data - this is used to create your model
- Configure your model with information about your data, what you want the model to do and some initial algorithm choices
- Build your model using the prepared data
- Iteratively test, review and revise the model until your are happy with the results
- Put your model into production, subscribe to your input data to target your ML Model
Business Rules
The business rules engine or expert system allows you to run event-driven actions based on the rules and configuration you make.
Business rules can be used to automate tasks based on changes that happen to any data in OpenDataDSL, such as:
- Run a quality check
- Derive some new data based on the changes made
- Add additional properties to updated data
Other Insights
We are continually working on adding more smart features to our existing applications as well as create new applications with smart features.
As most of the value in our data is within TimeSeries and Forward Curves, we are currently building some insights that will help you:
- Detect anomolies and alert you in real-time
- Analyse trends to give you a head start on potential up-coming events