Summary

Time series data poses new challenges and complexities. The chapter began with an introduction to important and popular datasets. We looked at different time series and their intricacies. Visualization of time series provides great insight, and the time series plots, along with the seasonal plot, are complementarily used for clear ideas and niche implementations. Accuracy metrics are different for the time series, and we looked at more than a handful of these. The concepts of ACF and PACF are vital in model identification, and seasonal components are also important to the modeling of time series. We also saw that different models express different datasets, and the degree of variation is something similar to the usual regression problems. The bagging of time series (ets only) reduces the variance of the forecasts. Combining heterogeneous base learners was discussed in the concluding section. The next chapter is the concluding chapter. We will summarize the main takeaways from the...