TAILIEUCHUNG - Báo cáo khoa học: "Parametric Models of Linguistic Count Data"

It is well known that occurrence counts of words in documents are often modeled poorly by standard distributions like the binomial or Poisson. Observed counts vary more than simple models predict, prompting the use of overdispersed models like Gamma-Poisson or Beta-binomial mixtures as robust alternatives. Another deficiency of standard models is due to the fact that most words never occur in a given document, resulting in large amounts of zero counts. We propose using zeroinflated models for dealing with this, and evaluate competing models on a Naive Bayes text classification task. Simple zero-inflated models can account for practically relevant. | Parametric Models of Linguistic Count Data Martin Jansche Department of Linguistics The Ohio State University Columbus OH 43210 USA jansche@ Abstract It is well known that occurrence counts of words in documents are often modeled poorly by standard distributions like the binomial or Poisson. Observed counts vary more than simple models predict prompting the use of overdispersed models like Gamma-Poisson or Beta-binomial mixtures as robust alternatives. Another deficiency of standard models is due to the fact that most words never occur in a given document resulting in large amounts of zero counts. We propose using zero-inflated models for dealing with this and evaluate competing models on a Naive Bayes text classification task. Simple zero-inflated models can account for practically relevant variation and can be easier to work with than overdispersed models. 1 Introduction Linguistic count data often violate the simplistic assumptions of standard probability models like the binomial or Poisson distribution. In particular the inadequacy of the Poisson distribution for modeling word token frequency is well known and robust alternatives have been proposed Mosteller and Wallace 1984 Church and Gale 1995 . In the case of the Poisson a commonly used robust alternative is the negative binomial distribution Pawitan 2001 which has the ability to capture extra-Poisson variation in the data in other words it is overdispersed compared with the Poisson. When a small set of parameters controls all properties of the distribution it is important to have enough parameters to model the relevant aspects of one s data. Simple models like the Poisson or binomial do not have enough parameters for many realistic applications and we suspect that the same might be true of log-linear models. When applying robust models like the negative binomial to linguistic count data like word occurrences in documents it is natural to ask to what extent the extra-Poisson variation has been .

• Báo cáo khoa học
• Parametric Models
• Linguistic Count Data
• Martin Jansche
• báo cáo khoa học
• mô hình ngôn ngữ
• xử lý ngôn ngữ tự nhiên
• Vitamin E
• Monitoring walnut
• Based on parametric
• Non parametric regression Models
• Parametric regression
• Technical Efficiency
• Phương pháp phân tích bao số liệu
• Hiệu quả kĩ thuật
• Stochastic frontier models
• Parametric approach
• Sheetmetal Design using Creo Parametric 2
• Sheetmetal Model Fundamentals
• Creating Primary Sheetmetal Wall Features
• Creating Secondary Sheetmetal Wall Features
• Bending and Unbending Sheetmetal Models
• Modifying Sheetmetal Models
• Apple production
• Trend analysis
• Nonparametric regression
• Apple (Malus domestica)
• Productivity and area trends
• Parametric regression model
• Auto regressive integrated moving average
• Normality test
• Jute production in west
• Time series parametric regression models
• Nonparametric Models
• Hypothesis Tests
• Unbiased Estimation
• Euclidean space
• Ebook Methods in human growth research
• Methods in human growth research
• Human growth
• Parametric models for postnatal growth
• Non linear longitudinal growth models
• Multilevel modeling
• intelligence and shrewdness
• proved tactless
• young and aspiring village
• literary and scientific
• authorized publication
• Medical Physicists
• báo cáo hóa học
• công trình nghiên cứu về hóa học
• tài liệu về hóa học
• cách trình bày báo cáo
• Vietnamese speech synthesis
• Context dependent
• Speech parameterization
• Statistical parametric speech synthesis
• Hidden Markov Models
• Mel frequency cepstral coefficient
• Fundamental frequency
• Bank efficiency
• Directional distance function
• Semi parametric estimation
• Vietnam’s and ASEAN banks
• Annals of Agricultural Sciences
• Agricultural Sciences
• Khoa học nông nghiệp
• Parametric stability
• Nonparametric stability
• GGE biplot
• Mathematical statistics
• Probability theory
• Fundamentals of statistics
• Estimation in parametric models