内容简介

Today's Web-enabled deluge of electronic data calls for automated methods of data analysis. Machine learning provides these, developing methods that can automatically detect patterns in data and then use the uncovered patterns to predict future data. This textbook offers a comprehensive and self-contained introduction to the field of machine learning, a unified, probabilistic approach. The coverage combines breadth and depth, offering necessary background material on such topics as probability, optimization, and linear algebra as well as discussion of recent developments in the field, including conditional random fields, L1 regularization, and deep learning. The book is written in an informal, accessible style, complete with pseudo-code for the most important algorithms. All topics are copiously illustrated with color images and worked examples drawn from such application domains as biology, text processing, computer vision, and robotics. Rather than providing a cookbook of different heuristic methods, the book stresses a principled model-based approach, often using the language of graphical models to specify models in a concise and intuitive way. Almost all the models described have been implemented in a MATLAB software package--PMTK (probabilistic modeling toolkit)--that is freely available online. The book is suitable for upper-level undergraduates with an introductory-level college math background and beginning graduate students.

作品目录

Chapter 1: Introduction
Chapter 2: Probability
Chapter 3: Statistics
Chapter 4: Gaussian models
Chapter 5: Generative models for classification
Chapter 6: Discriminative linear models
Chapter 7: Graphical Models
Chapter 8: Decision theory
Chapter 9: Mixture models and the EM algorithm
Chapter 10: Latent Linear models
Chapter 11: Hierarchical Bayes
Chapter 12: Sparce Linear Models
Chapter 13: Kernels
Chapter 14: Gaussian processes
Chapter 15: Adaptive basis function models
Chapter 16: Markov and hidden Markov Models
Chapter 17: State space models
Chapter 18: Conditional random fields
Chapter 19: Exact inference algorithms for graphical models
Chapter 20: Mean field inference algorithms
Chapter 21: Other variational inference algorithms
Chapter 22: Monte Carlo inference algorithms
Chapter 23: MCMC inference algorithms
Chapter 24: Clustering
Chapter 25: Graphical model structure learning
Chapter 26: Two-layer latent variable models
Chapter 27: Deep learning
· · · · · ·

作者简介

Kevin P. Murphy is Associate Professor in the Department of Computer Science and in the Department of Statistics at the University of British Columbia.

精彩摘录

Tounderstandtheseterms,youfirstneedtounderstandtheconceptoflikelihood.Assumeyouhaveaprobabilitydistribution-orratherfamilyofsuchdistributions-p(x;w)whichassignsaprobabilitytoeachdatapointx,givenaspecificsettingofitsparametersw.Thatis,differentvaluesoftheparameters,w,willchangetheprobabilityassignedtoeachdatapoint,x.Now,sincedifferentparameterscorrespondtodifferentdistributions,wecantunetheparametersinsuchawaythatthedatathatweobserve,D,isassignedahighprobabilityandpossibledatathatwedon'tobserveisassignedalowprobability.TothisendwedefinethelikelihoodfunctionL(D;w)=product_{xinD}p(x;w).Thatisthelikelihoodisjustthejointprobabilityoftheobserveddataasafunctionof...

——引自第71页

Inparticular,wedefinemachinelearningasasetofmethodsthatcanautomaticallydetectpatternsindata,andthenusetheuncoveredpatternstopredictfuturedata,ortoperformotherkindsofdecisionmakingunderuncertainty

——引自第1页