Moving Averages for Time Series Data Analysis

Let’s look at how we can use moving averages to spot trends in time-series data.

Time series are one of the most common datasets that you will come across in your Data Analytics career. A time series data-set is a collection of sequential data that is recorded in time intervals.

Some of the many examples include:

• The price of a stock over a certain period of time.
• Your heart rate over the course of a run measured by a fitness tracker.
• Records of the daily number of  visitors to a ski slope.

One of the main challenges when working with a time series is that there can be strong fluctuations, or noise, in the data. This can make it difficult to spot trends and key features that would otherwise be hidden under these fluctuations.

In this video we use a simple moving average to explore the trend in a time series of historical business revenue using MS Excel. We then go on to use this moving average as a baseline to measure monthly revenue against.

You can find the data-set on Kaggle (some columns & rows were deleted in this example to simplify things): https://www.kaggle.com/podsyp/time-series-starter-dataset