Probabilistic programming in practice
Probabilistic programming is a new way to create probabilistic models that allow you to predict or deduce new facts that are not in the results of observations. This allows, for example, to predict future events such as sales trends, computer system failures, experimental outcomes, and much more. The book is an introduction to probabilistic programming for practical programmers. The author almost immediately proceeds to practical examples: building a spam filter, diagnosing errors in the computer system, restoring digital images. You will get acquainted with probabilistic inference, where algorithms help predict, for example, the use of social networks. Along the way, you will learn about the use of a functional programming style for text analysis, object-oriented models for predicting the spread of tweets, and models with an open universe for measuring phenomena taking place in a social network. The book also contains chapters on how probabilistic models help in decision-making and modeling dynamic systems. The data you collect about customers, products and users of the site can help not only in interpreting the past, but also in predicting the future! Summary: - introduction to probabilistic modeling; - writing probabilistic programs on Figaro; - building Bayesian networks; - product lifecycle forecasting; - decision-making algorithms.
No reviews found
