▶What You Will Learn
- Visualize time series data and derive better insights
- Explore auto-correlation and master statistical techniques
- Use time series analysis tools from the stats, TSstudio, and forecast packages
- Explore and identify seasonal and correlation patterns
- Work with different time series formats in R
- Explore time series models such as ARIMA, Holt-Winters, and more
- Evaluate high-performance forecasting solutions
▶Key Features
- Perform time series analysis and forecasting using R packages such as Forecast and h2o
- Develop models and find patterns to create visualizations using the TSstudio and plotly packages
- Master statistics and implement time-series methods using examples mentioned
▶Who This Book Is For
Hands-On Time Series Analysis with R is ideal for data analysts, data scientists, and all R developers who are looking to perform time series analysis to predict outcomes effectively. A basic knowledge of statistics is required; some knowledge in R is expected, but not mandatory.
▶What this book covers
- Chapter 1, Introduction to Time Series Analysis and R, provides a brief introduction to the time series analysis process and defines the attributes and characteristics of time series data. In addition, the chapter provides a brief introduction to R for readers with no prior knowledge of R. This includes the mathematical and logical operators, loading data from multiple sources (such as flat files and APIs), installing packages, and so on.
- Chapter 2, Working with Date and Time Objects, focuses on the main date and time classes in R—the Date and POSIXct/lt classes—and their attributes. This includes ways to reformat date and time objects with the base and lubridate packages.
- Chapter 3, The Time Series Object, focuses on the ts class, an R core class for time series data. This chapter dives deep into the attributes of the ts class, methods for creating and manipulating ts objects using tools from the stats package, and data visualization applications with the TSstudio and dygraphs packages.
- Chapter 4, Working with zoo and xts Objects, covers the applications of the zoo and xts classes, an advanced format for time series data. This chapter focuses on the attributes of the zoo and xts classes and the preprocessing and data visualization tools from the zoo and xts packages
- Chapter 5, Decomposition of Time Series Data, focuses on decomposing time series data down to its structural patterns—the trend, seasonal, cycle, and random components. Starting with the moving average function, the chapter explains how to use the function for smoothing, and then focuses on decomposing a time series to down its components with the moving average.
- Chapter 6, Seasonality Analysis, explains approaches and methods for exploring and revealing seasonal patterns in time series data. This includes the use of summary statistics, along with data visualization tools from the forecast, TSstudio, and ggplot2 packages.
- Chapter 7, Correlation Analysis, focuses on methods and techniques for analyzing the relationship between time series data and its lags or other series. This chapter provides a general background for correlation analysis, and introduces statistical methods and data visualization tools for measuring the correlation between time series and its lags or between multiple time series.
- Chapter 8, Forecasting Strategies, introduces approaches, strategies, and tools for building time series forecasting models. This chapter covers different training strategies, different error metrics, benchmarking, and evaluation methods for forecasting models.
- Chapter 9, Forecasting with Linear Regression, dives into the forecasting applications of the linear regression model. This chapter explains how to model the different components of a series with linear regression by creating new features from the series. In addition, this chapter covers the advanced modeling of structural breaks, outliers, holidays, and time series with multiple seasonality.
- Chapter 10, Forecasting with Exponential Smoothing Models, focuses on forecasting time series data with exponential smoothing functions. This chapter explains the usage of smoothing parameters to forecast time series data. This includes simplistic models such as simple exponential smoothing, which is for time series with neither trend nor seasonal components, to advanced smoothing models such as Holt-Winters forecasting, which is for forecasting time series with both trend and seasonal components.
- Chapter 11, Forecasting with ARIMA Models, covers the ARIMA family of forecasting models. This chapter introduces the different types of ARIMA models—the autoregressive (AR), moving average (MA), ARMA, ARIMA, and seasonal ARIMA (SARIMA) models. In addition, the chapter focuses on methods and approaches to identify, tune, and optimize ARIMA models with both autocorrelation and partial correlation functions using applications from the stats and forecast packages.
- Chapter 12, Forecasting with Machine Learning Models, focuses on methods and approaches for forecasting time series data with machine learning models with the h2o package. This chapter explains the different steps of modeling time series data with machine learning models. This includes feature engineering, training and tuning approaches, evaluation, and benchmarking a forecasting model's performance.
