Time Series Model

Time Series takes a sequence of values and from them establishes forecasted values over a set horizon.

Select a DSD to use for training. Only those with a purpose of “Machine Learning” will be shown.

Select the Data Group within the DSD that contains the training data.

Three further entries are required:

• Moment – the field that represents the time in the sequence
• Index – the field that represents a numerical index in the sequence
• Value – the field that is the value to be forecast

If the model has been trained before, the right hand side of the page shows the metrics of that training. See below for more information on metrics.

Save the training settings and the click on “Train”

For Time Series training, there are a number of settings.

• Window Size - is set to the time period represented in the data cycle – e.g. 7
• Series Length - specifies the number of data points that are used when performing a forecast – e.g. 30
• Train Size - specifies the total number of data points in the input time series, starting from the beginning – e.g. 365
• Horizon – the number of values to forecast – e.g.7
• Confidence Level - indicates the percentage likelihood the real observed value will fall within the specified interval bounds – 95
• Training Data Cut-Off Index - specifies the Index field value that divides the training data from the test data. Ideally about half way through the data set.

Click Start to begin the training.

On completion metrics for the training will be shown.

• MAE – the average size of mistakes in predictions – the closer to zero the better
• RMSE – difference between predictions and actual - the closer to zero the better

There is no test option for Time Series