getInputSize

Returns the default input layer size for the neural network.

This method provides the standard input dimensionality used when no explicit input size is specified during model creation.

Default Configuration:

• Input Size: 3 neurons/features

• Rationale: Suitable for simple 3D feature vectors or RGB color channels

• Use Cases: Prototype development, educational examples, minimal viable networks

Feature Vector Examples:

[feature1, feature2, feature3] → Input Layer [3 neurons]
[x_coordinate, y_coordinate, z_coordinate] → Spatial data
[red_channel, green_channel, blue_channel] → Color analysis
[price, volume, volatility] → Financial indicators

Integration Notes:

• Value matches default parameter in createModel for consistency

• Can be overridden per model instance as needed

• Serves as baseline for architecture scaling decisions

Performance Impact:

• Memory: Minimal impact with only 3 input connections per first hidden neuron

• Computation: Fast forward pass suitable for real-time applications

• Training: Quick convergence for linearly separable problems

See also