uses moving window with overlap.
By the way, there is the method "without overlap".
"Amount of sales by month" and "Average temperture by month" use the edge of window the first day of the month. 0-1 Data Analysis is an application of this way.
We can also put the timing of the events as the edge of window. Analysis of Type 2 in Sensor Data Analysis is an application of this way.
The value made by "with overlap" is difficult to understand the meaning.
But the value made by "without overlap" is easy to understand the meaning because the edge of window has meanings.
And the value made by "without overlap" is easy to link other values made by same window. If month is the window, we can link "Amount of sales by month" and "Average temperture by month"
We can calculate easily average, standard deviation, maximum, minimum, correlation coeficient, etc. by general statistical software. But information we get through this method is limited.
But the values calculated by the window without overlap give us more information.
NEXT Sensor Data AnalysisTweet