A time series database (TSDB) is a software system that is optimized for handling time series data, arrays of numbers indexed by time (a datetime or a datetime range).
Any component that implements our data source Python interfaces. This can be a supported NoSQL database, a CSV file, or anything else that contains TSDB data. Data sources can contain information in arbitrary tables or documents. They can be queried using a time range to return timestamp and data relevant to the query.
Training is the process of converting history data into a machine learning model. The setting, features, and operations will vary based on the type of model used. Training is CPU (Or even GPU) intensive and data hungry. Training on time series data, with 10,000 aggregated data points will require between a few seconds to minutes on a common CPU. Also see model and inference.
A machine learning model uses features to represent changes in the data. With Loud ML, these features are assigned by the user when creating the model. For example, a feature can be
avg(cpu_load)to represent the average metric calculed on the document field named
cpu_load. The features are defined at model creation time and used both in training and inference.
This is the process of repeating the operations that have been discovered through training, this time using brand new data. For example, with time series data, running inference means your model will predict future data based on present and past data: if your features are
bucket_intervalis 60s you can predict the temperature and load in the next minute. You can run inference both using past history data, and present data.
A type used to represent the type of machine learning model you want to manipulate, e.g. a
donuttype if using the default unsupervised model provided in Loud ML.