Model-based Training System

A Model-based Training System is a training system that applies a model-based training algorithm to solve a model-based training task.

• AKA: Data-Driven Model Generator.
• Context:
  • It can range from being a Supervised Model-based Learning System to being an Unsupervised Model-based Learning System.
  • It can range from being a Supervised Model-based Classification System to being a Supervised Model-based Ranking System to being a Supervised Model-based Estimation System.
  • It can range from being a Lazy Model-based Learning System to being an Eager Model-based Learning System.
  • It can range from being a Discriminative Learning System to being a Generative Learning System.
  • It can be supported by a Model Training Library, or a Model Training Platform.
• Example(s):
  • a Regression System, such as a least-squares function fitting system.
  • a Statistical Modeling System, such as s Linear Model Learning system or a Logistic Regression Model Learning system.
  • a Decision Tree Learning System, such as a C4.5 system.
  • a Discriminative Machine Learning System, such as a logistic regression system.
  • a Neural-Network Learning System, such as a backprop learning system.
  • a Kernel-based Learning System, such as an SVM algorithm.
  • …
• Counter-Example(s):
  • an Instance-based Learning System, such as: k-Nearest Neighbor System.
  • a Locally Weighted Regression System.
• See: HMM System, CRF System.

---

---