So far in this session, you learnt about very simple forecasting models and also learnt how to evaluate the forecasts of these models on air passenger traffic data using popular error measures. In this segment, you will learn about another simple but popular forecasting method which is similar to the simple average method but forecasts better than it. This method is known as the **Simple Moving Average**.

Considering that the last observation in the time series has more impact on the future rather than the first observation, in the simple moving average method, we take the average of only the last few observations to forecast the future. Let’s understand more about it from Chiranjoy.

Now that you have an understanding of the Moving Average technique, let’s apply this technique to the air passenger traffic data to forecast the future.

Now that you have built the moving average model on the air passenger traffic data, let’s learn to calculate the RMSE and MAPE for the forecast.

Now, that you have learnt to forecast the future using the simple moving average technique, let’s understand what happens to the forecast if we change the moving average window for the air passenger traffic data.

In this segment, you learnt about moving average technique. In this technique, we consider the last few observations to forecast for the future. What if we assign some weights to those observations and then use them to forecast for the future? This is where the next set of forecasting techniques come into picture which you will learn in the upcoming segments.

But before that, let’s test the concepts you have learnt so far.

**Comprehension**

Now that you have a good understanding of the simple moving average method, let’s implement those learnings on another real-life dataset. You have been provided with the exchange rate dataset to build a simple moving average model on it. Exchange rate data consist of two columns, ‘Month’ and ‘Exchange rate’. The dataset can be downloaded from the link below.