Concepts / Approximation Error

Approximation Error

A hypothesis class is the set of possible classifiers from which the learning algorithm chooses.

  • Programming

The Choice Behind Learning

A learning algorithm does not choose from every classifier that could possibly exist. It chooses from a hypothesis class: the set of possible classifiers made available to the algorithm. That design choice creates a trade-off. If the class is too small, it may be unable to represent the classifier needed for the problem. If it is too rich, the available data may not be sufficient to choose reliably among all its possibilities.

The richest possible hypothesis class is not automatically the best one. A useful class must be flexible enough to avoid a large approximation error while remaining manageable enough to keep estimation error reasonable.

Two Sources of Difficulty

not contained by classtoo restrictedlimited datacomplexityOptimal classifierApproximation errorClass cannot contain theoptimal classifierEstimation errorSample and class complexityaffect reliable choiceChosen hypothesisclassFinite sample
Which part of the learning difficulty comes from the hypothesis class being unable to represent the target, and which part comes from learning with limited data?

Approximation error comes from the hypothesis class itself. It occurs when the class is not rich enough to contain the optimal classifier. Even if the learning algorithm had ideal information for choosing within that class, the desired classifier would still be unavailable.

Estimation error comes from having to learn with a finite sample and from the complexity of the hypothesis class. A richer class offers more possible classifiers, so the available sample may not be sufficient to choose reliably among them. In the source discussion, this difficulty is associated with overfitting.

A Class That Is Too Simple

not contained in classcannot representOptimal classifierNeeded for the problemUnrepresented targetpatternApproximation errorAvailableclassifiersSmall hypothesis class
What target patterns cannot be represented when the learner is restricted to a simple hypothesis class?

When flexibility is missing

Imagine a classification problem whose optimal classifier is not included in a small hypothesis class.

Restrict the choices: The learning algorithm can select only from the classifiers in the small class.

Search within the class: The algorithm may choose the best available classifier, but the optimal classifier remains unavailable.

Identify the remaining error: Because the class cannot contain the optimal classifier, the class restriction creates approximation error.

A small hypothesis class can reduce estimation error while increasing approximation error.

The important point is that more data alone does not change what the class can represent. If the optimal classifier is absent from the class, the representation limitation remains. This is why an overly restricted class may lead to underfitting.

A Class That Is Too Rich

many ways to fitfit observed datamay not generalizeFinite sampleObserved training dataTraining-data fitRich hypothesisclassMany possible classifiersPoor generalizationEstimation error
How does a very rich hypothesis class allow many possible classifiers, including classifiers that fit training data but generalize poorly?

When flexibility creates uncertainty

Imagine expanding a hypothesis class so that it contains many additional classifiers.

Add possible classifiers: The learner gains more flexibility and is less likely to be blocked by the class when searching for a good solution.

Increase the number of plausible fits: The same flexibility creates more ways to fit the particular finite sample.

Check the trade-off: The richer class may lower approximation error, but estimation error might increase if the sample is not sufficient to choose reliably.

A rich hypothesis class can reduce approximation error while increasing estimation error and the risk of overfitting.

Complexity Trade-Off

may causetends towardtends towardmight causeaims to balanceaims to balanceSmall classFew classifiersHigher approximationerrorPossible underfittingLower estimationerrorFewer choices todistinguishBalanced classReasonable flexibilityRich classMany classifiersLower approximationerrorMore target classifiersincludedPossible higherestimation errorPossible overfitting
What changes in approximation error and estimation error as the hypothesis class becomes more complex?
Hypothesis-class choiceApproximation errorEstimation errorTypical risk
Very smallMay increase because the optimal classifier may be absentTends to decrease because there are fewer choicesUnderfitting
Very richTends to decrease because more classifiers are availableMay increase because finite data may not distinguish the choices reliablyOverfitting
Reasonably designedNot excessively highReasonableA practical balance

Increasing class richness gives the learner more possible classifiers. Approximation error therefore decreases because the class is less likely to block a good solution. At the same time, estimation error might increase because it depends on both finite sample size and hypothesis-class complexity. Decreasing richness produces the reverse pressure.

Using Prior Knowledge

Choosing a hypothesis class is not only a contest between the smallest and richest possible classes. Prior knowledge about the problem can guide which classifiers should be included. Even a reasonable conjecture about the structure of the problem can help produce a class whose approximation error is not excessively high while its estimation error remains reasonable.

A domain-informed class

Suppose prior knowledge suggests that the classification problem has a particular structure, but the optimal classifier is unknown.

Use the available knowledge: Design a hypothesis class that includes classifiers consistent with the suspected structure.

Avoid unnecessary alternatives: Do not automatically include every possible classifier merely because a richer class might contain the optimum.

Balance the errors: Evaluate the design by asking whether it avoids excessive approximation error while keeping estimation error reasonable.

Prior knowledge helps define a practical class without requiring a complete description of the optimal classifier.

Practice Check

MEDIUM

A learner moves from a small hypothesis class to a richer one. State the likely direction of change for approximation error and the possible direction of change for estimation error. Then explain why neither a very small nor a very rich class is automatically best.

Hints
  • Ask first whether the richer class contains more possible classifiers.
  • Separate the class's ability to represent the optimal classifier from the finite sample's ability to support a reliable choice.
  • Use the terms underfitting and overfitting only after explaining the underlying trade-off.

What do you think happens?

If a class becomes richer, which error is most directly expected to decrease, and which error might increase?

  • Approximation error decreases; estimation error might increase
  • Approximation error increases; estimation error decreases
  • Both errors must decrease
  • Neither error can change
Reveal answer

Answer: Approximation error decreases; estimation error might increase

A richer class is less likely to exclude the optimal classifier, but its additional choices can make reliable learning from a finite sample more difficult.

Key Takeaways

  1. A hypothesis class is the set of possible classifiers from which the learning algorithm chooses.
  2. Approximation error occurs when the class is not rich enough to contain the optimal classifier.
  3. Estimation error is related to finite sample size and hypothesis-class complexity.
  4. A small class tends to reduce estimation error but may increase approximation error; a rich class tends to reduce approximation error but may increase estimation error.
  5. Prior knowledge can guide the design of a class that balances representational flexibility with reliable learning.

Key Takeaways

  • Approximation error is caused by a hypothesis class that cannot represent the optimal classifier.
  • Estimation error is connected to finite sample size and the complexity of the hypothesis class.
  • Increasing class richness usually lowers approximation error but may raise estimation error.
  • A practical hypothesis class should balance flexibility with reliable learning.
  • Prior knowledge can help select a useful class even when the optimal classifier is unknown.