Normalising Inputs
Agency - along with some other ARDI products - uses a Normalisation Engine to convert the inputs you're getting into inputs suitable for feeding an AI.
In most cases, AIs need normalised inputs, meaning that they all have a similar numeric range.
Otherwise, large values (such as pressures, which often go over 1000kPa) would have a much larger impact on your AI than small values (such as ambient temperature, which range from 0-50). Instead, these values are both reduced to simple 0-1 values.
When creating an agent, you can specify the 'normalisation format' you'd like to use. At a very basic level, you can use the following…
| Format | Meaning |
|---|---|
| A | All inputs are automatically scaled between 0-1, except for those that cross between negative and positive numbers, which are scaled -1 to 1. Values that range between negative numbers and zero are inverted. |
| -1-1 | All inputs are scaled to fit into a scale between -1 and +1 |
| 0-1 | All inputs are scaled to fit in a range between 0-1 |
| * | Inputs are not scaled and left as raw numbers |
For more information, you might want to see the Normalisation Engine documentation.
Note that the normalisation is performed before the agent is executed, so if you're normalising your values you'll need to ensure any parameters you provide the Agent are also normalised.
For legacy reasons, -1-1 is the default normalisation method.