Observation Language

An Observation Language is the language that can describe observations which are represented as an input set used by a machine learning system.

• AKA: Instance language.
• Example(s):
  • a fixed set of attributes.
  • …
• Counter-Example(s):
  • Hypothesis Language (output).
• See: First-Order Logic; Hypothesis Space; Inductive Logic Programming; Propositional Language; Instance-Based Learning; Clustering; Neural Network; Decision Tree; Rule Set; Graphical Model, Attribute-Value Learning.

---

---

References

2017

• (Blockeel, 2017) ⇒ Hendrik Blockeel. (2017). "Observation Language" In: (Sammut & Webb, 2017).
  • QUOTE: The observation language used by a machine learning system is the language in which the observations it learns from are described.

    (...) Most machine learning algorithms can be seen as a procedure for deriving one or more hypotheses from a set of observations. Both the input (the observations) and the output (the hypotheses) need to be described in some particular language and this language is called the observation language or the Hypothesis Language respectively. These terms are mostly used in the context of symbolic learning, where these languages are often more complex than in subsymbolic or statistical learning.

    (...) Probably the most used setting in machine learning is the attribute-value setting (see Attribute-Value Learning). Here, an example (observation) is described by a fixed set of attributes, each of which is given a value from the domain of the attribute. Such an observation is often called a vector or, in relational database terminology, a tuple. The attributes are usually atomic (i.e., not decomposable in component values) and single-valued (i.e., an attribute has only one value, not a set of values).

---