Local Distance Metric Adaptation Algorithm

A Local Distance Metric Adaptation Algorithm is a kernel learning algorithm that can tune every local kernel's distance metric.

• AKA: Kernel Shaping, Nonstationary Kernels, Supersmoothing.
• Example(s):
  • …
• Counter-Example(s):
  • a Gaussian Process Algorithm,
  • a k-Nearest Neighbor Learning Algorithm,
  • a Maximum Entropy-based Learning Algorithm,
  • a Support Vector Machine Learning Algorithm.
• See: Locally Weighted Regression for Control, Locally Weighted Regression Algorithm, Learning Control, Adaptive Control.

---

---

References

2017

• (Sammut & Webb, 2017) ⇒ (2017). "Local Distance Metric Adaptation”. In: (Sammut & Webb, 2017)
  • QUOTE: In learning systems with kernels, the shape and size of a kernel plays a critical role for accuracy and generalization. Most kernels have a distance metric parameter, which determines the size and shape of the kernel in the sense of a Mahalanobis distance. Advanced kernel learning tune every kernel’s distance metric individually, instead of turning one global distance metric for all kernels.

---