multinomial

It didn't have anything built-in, but so we wrote a few functions for multinomial logistic regression, and that was good enough.

For better accuracy with small samples you could use the multinomial distribution instead.

It is not extremely difficult to build a multinomial regression model and find "adversarial" images for this trained model.

See the discussion elsewhere in this thread for confidence intervals in multinomial distributions.

Absolutely, data++in fact one of last year's fellows did multinomial mixture modeling and boosted alternating decision trees to, as he put it, improve okcupid's "love machine".

Propensity scores were estimated via multinomial logistic regression of treatment group.

Re: "Empirical studies showed that the mutli-variate Bernoulli model is inferior to the multinomial Bayes model, and the latter has been shown to reduce the error by 27% on average [13].

Then they mention some common distributions, especially Gaussian, exponential, Poisson, multinomial, and uniform.

I rephrased it a little bit to"Empirical comparisons provide evidence that the multinomial model tends to outperform the multi-variate Bernoulli model if the vocabulary size is relatively large [13].

In practice, it is recommended that the choice between a multi-variate Bernoulli or multinomial model for text classification should precede comparative studies including different combinations of feature extraction and selection steps.

In this form, one of the sufficient statistics often has an interpretation as a "pseudo-count"; how many effective samples are encoded in your prior?exponential family distributions include: poisson, exponential, bernoulli, multinomial, gaussian, negative binomial, etc.

We find that the multi-variate\nBernoulli performs well with small vocabulary sizes,\nbut that the multinomial performs usually performs\neven better at larger vocabulary sizes—providing on\naverage a 27% reduction in error over the multi-variate\nBernoulli model at any vocabulary size.