Well-Posed Learning Problems



A Computer Program is said to be Learn from Experience E with respect to some class of Tasks T and Performance P , if its performance at tasks in T as Measured by P, improves with Experience E.

Any problem can be segregated as well-posed learning problem if it has three traits :

Task - T
Performance Measure - P
Experience - E

In General to have well defined learning problem we must identify three features the class of Tasks, the measure of performance to be improved, source of experience.

Examples that efficiently defines the well-posed learning problem are:

  1. A checkers learning problem
    Task – Playing checkers game - T
    Performance Measure – percent of games won against opposer -P
    Training Experience – playing implementation games against itself -E

  2. Handwriting Recognition Problem
    Task – Acknowledging handwritten words within portrayal -T
    Performance Measure – percent of words accurately classified -P
    Experience – a directory of handwritten words with given classifications -E

  3. A Robot Driving Problem
    Task – driving on public four-lane highways using sight scanners -T
    Performance Measure – average distance progressed before a fallacy -P
    Experience – order of images and steering instructions noted down while observing a human driver -E

  4. Fruit Prediction Problem
    Task – forecasting different fruits for recognition -T
    Performance Measure – able to predict maximum variety of fruits -P
    Experience – training machine with the largest datasets of fruits images -E

  5. Face Recognition Problem
    Task – predicting different types of faces -T
    Performance Measure – able to predict maximum types of faces -P
    Experience – training machine with maximum amount of datasets of different face Images -E

  6. Automatic Translation of documents
    Task – translating one type of language used in a document to other language -T
    Performance Measure – able to convert one language to other efficiently -P
    Experience – training machine with a large dataset of different types of languages-E

  7. To better filter emails as spam or not
    Task – Classifying emails as spam or not -T
    Performance Measure – The fraction of emails accurately classified as spam or not spam -P
    Experience – Observing you label emails as spam or not spam -E