Frequency Profile
The distributional shape describing how many times each distinct sample point appears in the training set.
Timeline (last 90 days)
As the sample size increases, under the constructed priors both the penalty quantity and the actual risk quantity tend to vanish, but their ratio can diverge exponentially.
Under the exchangeability assumption, the full-sample posterior probability depends only on the multiplicities of the samples and the complete frequency profile.
The corrected penalty term is no longer a linear function of a single frequency, but a function of the entire frequency profile.
The original frequency coefficient approximates well when the frequency profile distribution is concentrated, and deviates from the true value when the distribution is dispersed and rare profiles dominate.
All Facts (4)
As the sample size increases, under the constructed priors both the penalty quantity and the actual risk quantity tend to vanish, but their ratio can diverge exponentially.
50%UnverifiedUnder the exchangeability assumption, the full-sample posterior probability depends only on the multiplicities of the samples and the complete frequency profile.
50%UnverifiedThe corrected penalty term is no longer a linear function of a single frequency, but a function of the entire frequency profile.
50%UnverifiedThe original frequency coefficient approximates well when the frequency profile distribution is concentrated, and deviates from the true value when the distribution is dispersed and rare profiles dominate.
50%