Table of Contents

Class SoftLocallyWeightedOp

Namespace
AiDotNet.JitCompiler.IR.Operations
Assembly
AiDotNet.dll

Represents a soft locally-weighted regression operation for differentiable instance-based learning.

public class SoftLocallyWeightedOp : IROp
Inheritance
SoftLocallyWeightedOp
Inherited Members

Remarks

Implements differentiable locally-weighted regression using attention-based weighting. This enables gradient-based optimization and JIT compilation of LOESS/LOWESS-style models.

The operation computes: 

distances[i] = ||input - X_train[i]||²
weights = softmax(-distances / bandwidth)
output = Σ weights[i] * y_train[i]

For Beginners: Locally-weighted regression makes predictions by computing a weighted average of nearby training examples, where "nearby" is determined by distance.

This soft version uses attention (softmax) to compute weights, making it fully differentiable:

  • Points close to the query get high attention weights
  • Points far from the query get low attention weights
  • The bandwidth controls how quickly attention drops off with distance

Properties

Bandwidth

Gets or sets the bandwidth parameter controlling the locality of weighting. Smaller bandwidth = more local (only nearby points matter).

public double Bandwidth { get; set; }

Property Value

double

Methods

ToString()

Gets a string representation of this operation for debugging.

public override string ToString()

Returns

string

A string describing this operation.

Remarks

The string format is: "tOutput = OpType(tInput1, tInput2, ...) : Type [Shape]"

For Beginners: This creates a readable description of the operation.

Example outputs:

  • "t2 = Add(t0, t1) : Float32 [3, 4]"
  • "t5 = MatMul(t3, t4) : Float32 [128, 256]"
  • "t8 = ReLU(t7) : Float32 [32, 128]"

This is super helpful for debugging - you can see exactly what each operation does and what shape tensors flow through the graph.

Validate()

Validates that this operation is correctly formed.

public override bool Validate()

Returns

bool

True if valid, false otherwise.

Remarks

Basic validation checks that the operation has required information. Derived classes can override to add operation-specific validation.

For Beginners: This checks that the operation makes sense.

Basic checks:

  • Output ID is valid (non-negative)
  • Has the right number of inputs
  • Shapes are compatible

Specific operations add their own checks:

  • MatMul: inner dimensions must match
  • Conv2D: kernel size must be valid
  • Reshape: total elements must be preserved

If validation fails, the operation can't be compiled.

Edit this page