Simplest AI Trick in the Book Normalised Tunable Sigmoid Function Dino Dini NHTV University of Applied Sciences
Jul 08, 2015
Simplest AI Trick in the BookNormalised Tunable Sigmoid Function
Dino Dini NHTV University of Applied Sciences
Normalized Values Are Useful
For example:
● Utility calculations
● Input management
● Control systems
● Tunable parameters
Analog Input
Abstract away the device dependent positional values
(0 to 255? -1024 to 1024?) and normalise.
Normalised values are much easier to work with.
-1 10
Example:Analog Input - Left / Right Rotation
Input Device
1024
-1024
Movement Driver
1
-1
RotationDegrees per
Frame
Normalizer
5
-5
Example:Analog Input - Left / Right Rotation
Input Device
1024
-1024
Movement Driver
1
-1
RotationDegrees per
Frame
Normalizer
5
-5
Linear relationship
Degrees rotation per frame
Control input (left - right)
Linear relationship
Degrees rotation per frame
Control input (left - right)
I want greater sensitivity
Linear relationship
Degrees rotation per frame
Control input (left - right)
Linear relationship
Degrees rotation per frame
Control input (left - right)
I also want full range
Linear relationship
Degrees rotation per frame
Control input (left - right)
Greater sensitivity
Full Range
Example:Analog Input - Left / Right Rotation
Input Device
1024
-1024
Movement Driver
1
-1
RotationDegrees per
Frame
Normalizer
5
-5
Sigmoid like
function
1
-1
Example:Analog Input - Left / Right Rotation
Input Device
1024
-1024
Movement Driver
1
-1
RotationDegrees per
Frame
Normalizer
5
-5
Sigmoid like
function
1
-1
k
Sigmoid function?
Logit function?
Normalised Tunable (half) Sigmoid Function?
Normalised Tunable (half) Sigmoid Function?
k = 0.2
Normalised Tunable (half) Sigmoid Function?
k = 0.01
Normalised Tunable (half) Sigmoid Function?
k = 2
Normalised Tunable (half) Sigmoid Function?
k = -1.2
Normalised Tunable (half) Sigmoid Function?
k = -1.01
Normalised Tunable (half) Sigmoid Function?
k = -3
Normalised Tunable Sigmoid Function
k = 0.2
Normalised Tunable Sigmoid Function
Thank you