Data Science Course In Mumbai

Wednesday, August 07, 2019 9 Comments

Today`s discussion is all about Forecasting Methods based on Smoothing. This article shows the significance of forecasting methods which we use to forecast the future. Based on smoothing, now we will discuss forecasting methods. There are two forecasting methods based on smoothing. They are

Moving Average and
Exponential Average

Moving average:

In general moving average is an extension of Naive forecasting technique. The naive technique is a method which we use to forecast the future. So the technique works in this way

“Tomorrow`s forecasted value is equal to today`s forecast value”.

Moving average is all about taking past few observation under consideration and taking an average of that. The moving average will smooth the noise I mean performing moving average will handle random components or Noise in data. In short moving average will forecast future points by using an average of several past points. It is more suitable for specific series with no trend or no seasonality.

For example, a time slot of the MA reveals the level & Trend of the series. Filtering can be done in seasonal & random components. It actually helps us to remove seasonality and compute the seasonal index. There are two different calculations associated with moving average. They are

Centered moving average.
Trailing moving average

Centered moving average:

Here centered moving average calculates based on window centered on time ‘t’.

For suppose, if we calculate the moving average of any observation we need to consider equal observations based around time. So centered moving average consider an equal number of observations ahead of time and the same equal number of observations behind time.

The centered moving average is not a perfect suit to forecast the future. We mostly use the centered moving average for visualization purpose.

Trailing Moving average:

Trailing moving average calculates base on the window from time ‘t’ & backward.

Exponential Smoothing:

Forecasted value is equal to the current value and yesterday`s forecasted value. We also need to forecasted value and current value.

Finally, success depends on choosing window width and over (or) under smoothing would have a bad impact on smoothing.

This article explains your basics about Forecasting methods based on smoothing.

So the centered moving average, Trailing moving average and Exponential smoothing holds great significance playing its role in the space of data science. Each of the moving averages will play their roles in meeting the requirements.

data science course in mumbai

ExcelR is one of the leading training providers of professional certification training solutions in the world. We strive to provide the best training methods, taught by industry professionals at locations all over the world. Partner with us today and see why we are the number one choice for business professionals all over the world.