Forecasting: principles and practice

This book is an excellent resource for anyone trying to master practical nuts and bolts of forecasting or who is just starting to study the field. The authors explain the practical issues needed to forecast. If you want to know about the distribution of the Durbin-Watson statistic, or other recondite details, this is not the right resource. The text is tightly integrated with R examples which make it easy to start applying immediately what you have learned. Note: I read the free web version before the text was released. An index, however, would have been helpful.

While working on forecasting (understand “time series analysis”) I found several interesting and state of the art articles from Rob J. Hyndman. He is the co-author, with George Athanasopoulos of Forecasting: Principles and Practice. This is an excellent, concise and comprehensive text explaining concepts behind forecasting, common algorithms and how to implement them in R (for a business view of forecasting, I advise "Future Ready").The book presents key concepts of forecasting. From judgemental forecasting (which can be useful when you have no or few data) to simple/multiple regression, time series decomposition, exponential smoothing (ETS), ARIMA and a few more advanced topics such as Neural Networks. I would suggest to the author to add Support Vector Regression (SVR) and ensemble learning for the next edition of the book. Each concept of the book is covered through examples with real data. What is most appreciable about the book is how concise and readable it is. Each sentence is useful to understand the described concept, nothing superfluous.The book contains good overview and schema about each technique and how to set their meta-parameters. The R codes are well presented and easy to implement and test. The book can easily be used to teach forecasting since each chapter contains exercises. In conclusion, Forecasting: Principles and Practice is THE book to learn time series analysis algorithms and how to implement them in R.

Excellent book with very broad coverage. Depth may be lacking some times and you may have to resort to the academic papers cited. There is no coverage of recent deep learning models like RNN and LSTM for forecasting.

Great goob, starting from simple to the complex. Good reference for the data scientist.

Money well-spent. RH knows his stuff, and is a teacher so ostensibly cares about whether his students/readers actually learn something. I'm a huge fan of his work with the R package "forecast" and his various other offerings. You'd be wise to turn your attention to him.

This is a very good book for learning forecasting, with an emphasis on applying the ideas within the "R" environment.

Great reference book for time series forecasting

The pdf found everywhere on the web for free, The Little Book of R for Time Series, gives you everything you need. This book is mostly fluff.