Skip to content

Khinchin Theorem

  • For many sequences, the average of the logarithms of the terms converges to a fixed constant (the Khinchin constant).
  • This convergence can be used as a diagnostic for randomness in a sequence.
  • The theorem appears in probability theory and is applied to random processes, random number generation, and algorithm analysis.

Khinchin theorem is a mathematical result that states that, in a given sequence of numbers, the average of the logarithms of those numbers converges to a constant value. This constant value is known as the Khinchin constant and is denoted by the letter “K”.

The theorem asserts convergence of the mean of logarithms for terms of a sequence to a single constant (the Khinchin constant). Practically, this provides a quantitative reference: the closer the average of the logarithms of a sequence is to the Khinchin constant, the more the sequence is considered random according to the criterion described. The result is cited as important in probability theory and is used in the study of random processes, random number generation, and in analyses of algorithms.

The sequence: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10

nlog(n)
10
20.301
30.477
40.602
50.699
60.778
70.845
80.903
90.954
101

The average of these logarithms is 0.564, which is close to the Khinchin constant (K = 0.5685).

The sequence: 2, 3, 5, 7, 11, 13, 17, 19, 23, 29

primelog(prime)
20.301
30.477
50.699
70.845
111.039
131.127
171.230
191.278
231.360
291.440

The average of these logarithms is 0.838, which is also close to the Khinchin constant (K = 0.5685).

  • Study of random processes.
  • Assessment of random number generation quality.
  • Analysis and evaluation of algorithm performance using averages of input-size logarithms.
  • A key implication is that the theorem provides a way to measure the randomness of a sequence: the closer the average of the logarithms of a sequence is to the Khinchin constant, the more random the sequence is considered to be (per the statement in the source).
  • Khinchin constant (K = 0.5685)
  • Probability theory
  • Random processes
  • Random number generation
  • Analysis of algorithms