We've had quite an amazing and rewarding journey through the realm of machine learning. We have learned how to use algorithms to classify labeled data and apply our findings to make recommendations. We have seen how to extract business value from raw, unlabeled information by using unsupervised machine learning and clustering algorithms. However, one key component has been missing from our analysis so far—the temporal dimension.
Time is money, the saying goes—and as such, organisations of all sizes, from small businesses to large corporations, to governments, to complex multinational institutions such as the European Union, continuously measure and monitor a multitude of economic indicators over time. To be meaningful, the data is collected at regular intervals, allowing analysts to identify hidden structures and patterns, and predict future developments based on past and present conditions. These values, measured regularly on a time scale...