Data binning

Data binning, also called data discrete binning or data bucketing, is a <a href="/facts/Data_pre-processing/CS2W8m5B">data pre-processing</a> technique used to reduce the effects of minor <a href="/facts/Observation_error/q7pa7bRt">observation errors</a>. The original data values which fall into a given small interval, a <a href="/facts/Bin_(computational_geometry)/yoYpch6L">bin</a>, are replaced by a value representative of that interval, often a <a href="/facts/Central_tendency/9yhcQ6me">central value</a> (<a href="/facts/Mean/swcEd4Pg">mean</a> or <a href="/facts/Median/YwXGWTyr">median</a>). It is related to <a href="/facts/Quantization_(signal_processing)/fddS3BE2">quantization</a>: data binning operates on the <a href="/facts/Abscissa/TGAvIATE">abscissa</a> axis while quantization operates on the <a href="/facts/Ordinate/TGAvIATE">ordinate</a> axis. Binning is a generalization of <a href="/facts/Rounding/rCLYXN00">rounding</a>.
Statistical data binning is a way to group numbers of more-or-less continuous values into a smaller number of "bins". For example, if you have data about a group of people, you might want to arrange their ages into a smaller number of age intervals (for example, grouping every five years together). It can also be used in <a href="/facts/Multivariate_statistics/tDuNcWrL">multivariate statistics</a>, binning in several dimensions at once.
In <a href="/facts/Digital_image_processing/zenalNCf">digital image processing</a>, "binning" has a very different meaning. <a href="/facts/Pixel_binning/EFHuPf2o">Pixel binning</a> is the process of combining blocks of adjacent <a href="/facts/Pixel/UzwBvXBV">pixels</a> throughout an image, by summing or averaging their values, during or after readout. It reduces the amount of data; also the relative noise level in the result is lower.

Data binning open-in-new

Data binning