Structured Prediction Task

A Structured Prediction Task is a prediction task with either a complex input or a complex output or both.

• AKA: Structured-Output Learning Task.
• Context:
  • It can range from being a Structured Classification Task to being a Structured Ranking Task.
• Example(s):
  • Conditional Random Field,
  • Neural Structured Learning Task,
  • POS Tagging,
  • Structured Support Vector Machine.
  • Structured k-Nearest Neighbours.
  • …
• Counter-Example(s):
• a Simple Prediction Task, such as a Univariate Prediction Task.
• See: Structured Model Learning Algorithm, Classification Algorithm, Ranking Algorithm.

---

---

References

2020

• (Wikipedia, 2020) ⇒ https://en.wikipedia.org/wiki/Structured_prediction Retrieved:2020-2-28.
  • Structured prediction or structured (output) learning is an umbrella term for supervised machine learning techniques that involves predicting structured objects, rather than scalar discrete or real values.

    Similar to commonly used supervised learning techniques, structured prediction models are typically trained by means of observed data in which the true prediction value is used to adjust model parameters. Due to the complexity of the model and the interrelations of predicted variables the process of prediction using a trained model and of training itself is often computationally infeasible and approximate inference and learning methods are used.

2008

• (Elkan, 2008) ⇒ Charles Elkan. (2008). “Log-linear models and conditional random fields." Notes for a tutorial at CIKM-2008 (CIKM 2008).
  • QUOTE: POS tagging is an example of what is called a structured prediction task. The goal is to predict a complex label (a sequence of POS tags) for a complex input (an entire sentence). The word “structured” refers to the fact that labels have internal structure, in this case being sequences

---