What to expect?
In this course you will learn how to utilize the Machine Learning Trainer application. This application enables a user to create machine learning models with the data that you are collecting on SORBA. In the ML Trainer you can create algorithms for anomaly detection, optimization, and for other applications. These models allow you to analyze and leverage your data in order to take better control over the machines that you have connected to SORBA.
Lesson List
- ML Trainer Overview - This video goes over the basics of the machine learning trainer.
- Dataset Creation - In this video you will learn how to create a dataset. Datasets can be made using hot data from your assets or can be made with a historical csv file.
- Datasets - This video shows how to navigate your datasets. You will also learn how to access and use the statistical tools provided within the explore page for your datasets.
- Dataset Drift- This video goes over the concept of data drift and how you can use this calculation when you are creating a dataset.
- Projects - This video goes over the projects tab within the ML trainer. Here you can see the process for creating new projects and analyses.
- Clustering Models - Clustering models are used for anomaly detection solutions. Here you will learn how to create this type of model.
- Work Items - Work items provide a history of the different events experienced by your system. This video goes over how to create these items. Work items are necessary for the creation of classification algorithms.
- Work Items from Dashboard - This video goes over an alternate method for the creation of work items.
- Classification Models - Classification models allow you to classify and predict specific events. This video shows how to create this type of model.
- Regression Models - Regression models allow you to predict a single value within your system. This video teaches you how to create this type of model.
- Optimization Models - Optimization models allow you to maximize or minimize a variable within your system. This video teaches you how to create this type of model.
- Digital Twin Models - Digital twins allow you to generate a virtual recreation of your system where you can predict multiple values. This video teaches you how to create this type of model.
- Forecasting Models - Forecasting models allow you to predict future values. This video teaches you how to create this type of model.
- Offline Predictor - The offline predictor is a tool which allows you to simulate how your model would react to data before you deploy it into a real time environment. This is helpful to ensure the success of your model before it is actually deployed.
