different between degrade vs overlearn

English

Etymology

From Middle French dégrader

Pronunciation

• (UK) IPA^(key): /d????e?d/, /di???e?d/

Rhymes: -e?d

Verb

degrade (third-person singular simple present degrades, present participle degrading, simple past and past participle degraded)

1. (transitive) To lower in value or social position.
   • 1859-1890, John G. Palfrey, History of New England to the Revolutionary War

     Prynne was sentenced by the Star Chamber Court to be degraded from the bar.

2. (intransitive, ergative) To reduce in quality or purity.

3. (transitive, geology) To reduce in altitude or magnitude, as hills and mountains; to wear down.

Derived terms

• degradation

Translations

Portuguese

Verb

degrade

1. first-person singular present subjunctive of degradar
2. third-person singular present subjunctive of degradar
3. third-person singular imperative of degradar

Spanish

Verb

degrade

1. Formal second-person singular (usted) imperative form of degradar.
2. First-person singular (yo) present subjunctive form of degradar.
3. Formal second-person singular (usted) present subjunctive form of degradar.
4. Third-person singular (él, ella, also used with usted^?) present subjunctive form of degradar.

English

Alternative forms

• over-learn

Etymology

From over- +? learn.

Pronunciation

• (UK) IPA^(key): /??v??l??n/

Verb

overlearn (third-person singular simple present overlearns, present participle overlearning, simple past and past participle overlearned or overlearnt)

1. To learn (something) more than is necessary; to study excessively, to take (something) too much to heart.
   • 2010, "An own goal on gay rights", The Economist, 14 Oct 2010:

     Why the rigmarole of a military survey before enacting a measure most Americans already favoured? Maybe he over-learned the lesson of the clobbering Bill Clinton got when he barged into this area right at the start of his presidency.

2. (psychology, education) To learn (something) to the point where responses become instinctive.
3. (modeling) Mostly when talking about neural networks, to learn a task to the point where responses actually start to degrade. Compare with overfit, in model tuning contexts.

Translations