Time Series Analysis

In many situations, we need to explore and model how data changes over time, which allows us to make predictions about what will happen in the future. In particular, businesses are interested in forecasting future needs and demands. For example, a company that produces a certain product needs accurate predictions about their volume of sales in the near future in order to make decisions about things such as production schedules, the purchasing of raw materials, inventory policies, and sales quotas. A good forecast ensures the company is operating efficiently and cost-effectively. A poor forecast may result in poor planning and increased costs.

In time series analysis, data is collected over a period of time at regular intervals, such as monthly, quarterly, or yearly. Using this time-based data, a model can be built to predict what will happen in future time intervals. For example, a business might collect quarterly sales revenue over a three-year period and then use that data to build a model to predict the sales revenue for each quarter of the fourth year. There are several different time series models, but some models will give more accurate predictions than others depending on the traits in the particular time series data. When using time series models, we need to understand how the different models make their forecasts and identify which model to use on a given time series to ensure the most accurate predictions from the model.