Entropíkóðun

Entropíkóðun, or entropy coding, is a form of lossless data compression used in computer science and information theory. It is a type of statistical data compression that works by assigning shorter codes to more frequent symbols and longer codes to less frequent symbols. The theoretical limit for compression is determined by the entropy of the data source, which is a measure of its randomness or unpredictability.

There are several algorithms that fall under the umbrella of entropy coding, including Huffman coding and arithmetic coding. Huffman coding constructs a prefix code, where no code is a prefix of another, based on the probabilities of each symbol. Arithmetic coding represents an entire message as a single fraction, where the range of the fraction is subdivided according to the probabilities of the symbols.

The effectiveness of entropy coding depends on the accuracy of the probability model used. If the model accurately reflects the true probabilities of the symbols in the data, then high compression ratios can be achieved. However, if the model is inaccurate, the compression may be less effective or even result in an increase in data size.

Entropy coding is widely used in various applications, such as image and video compression (e.g., JPEG, MPEG), audio compression (e.g., MP3), and general-purpose file compression utilities (e.g., ZIP, GZIP). Its ability to exploit statistical redundancies in data makes it a fundamental technique in modern data processing and storage.