On deviations between kernel-type estimators of a distribution density in p ⩾ 2 independent samples

ABSTRACT

In the article, the tests are constructed for the hypotheses that p ⩾ 2 independent samples have the same distribution density (homogeneity hypothesis) or have the same well-defined distribution density (goodness-of-fit test). The limiting power of the constructed tests is found for some local “close” alternatives.

KEYWORDS:

• Goodness-of-fit test
• Histogram
• Homogeneity hypothesis
• Kernel-type estimator of density
• Power of test
• Test consistency
• Wiener process.

MATHEMATICS SUBJECT CLASSIFICATION:

• 62G10
• 62G20