Discovering linear trends
Perhaps the simplest trend is the linear trend. Naturally, we would only attempt to find such a trend on serial data, such as data ordered by time. For this example, we will use the daily frequency of email messages on the R-help mailing list (https://stat.ethz.ch/mailman/listinfo/r-help), an email list for users seeking help with the R programming language. The mailing list archive includes every message and the time it was sent. We wish to find a daily linear trend of message frequency, as opposed to hourly, minutely, monthly, yearly, and so on. We must decide the unit of frequency before applying trend or anomaly analysis, as the count of messages per day may be strongly linear while the count per hour may be non-linear ...
Get AI Blueprints now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
