Automatic taggers describe music in terms of a multinomial distribution over relevant semantic concepts. This paper presents a framework for improving automatic tagging of music content by modeling contextual relationships between these semantic concepts. The framework extends existing auto-tagging methods by adding a Dirichlet mixture to model the contextual co-occurrences between semantic mul...

We explore a model of stress prediction in Russian using a combination of local contextual features and linguisticallymotivated features associated with the word’s stem and suffix. We frame this as a ranking problem, where the objective is to rank the pronunciation with the correct stress above those with incorrect stress. We train our models using a simple Maximum Entropy ranking framework all...

This is not immediately obvious (for example, this ratio of factorials is not a product of multinomial coefficients) but it is not hard to prove. The only proof I know proceeds by checking that the valuations vp(un) are non-negative for every prime p; an interpretation of un as counting natural objects or being dimensions of natural vector spaces is far from clear. As it turns out, the generati...

In this paper, we develop new test statistics for private hypothesis testing. These statistics are designed specifically so that their asymptotic distributions, after accounting for noise added for privacy concerns, match the asymptotics of the classical (nonprivate) chi-square tests for testing if the multinomial data parameters lie in lower dimensional manifolds (examples include goodness of ...

GLM|LIC Working Paper No. 28 | April 2017 Childhood Determinants of Internal Youth Migration in Senegal* We analyze the childhood socio-economic determinants of youth decision to internally migrate in Senegal. Young people undertake mostly rural-to-rural and urban-to-urban migrations and over half of them are temporary migrants. Using multinomial logit models and a unique household survey in Se...

This paper describes two Stata programs, mclgen and mclest, for imposing special restrictions on multinomial logistic models. MCL stands for Multinomial Conditional Logit, a term coined by Breen (1994). An MCL model uses a conditional logit program to estimate a multinomial logistic model. This produces the same log likelihood, estimates and standard errors, but allows greater flexibility in im...

Article history: Available online xxxx