Math

 

Statistic

Figures of merits to extract information about a data set.

>>> Average and RMS Average<<<

Computationally and memory efficient definitons of: Average, RMS averag, Variance and Standard Deviation. 

Linear Regression

Linear Regression is a tool to estimate a linear relationship between an input set and an output set.

Those formulae allows computation of Gain, Bias and Error from just five accumulation variables. This eliminates the need of storing and processing all samples of the set (2N variabiles)

>>>Linear Regression<<< 

Bias and Gain

>>>Linear Regression: Derivation<<<

Compute linear regression parameters from the definition of the errror.

>>>Linear Regression Error<<<

Error Estimation

>>>Multidimensional Linear Regression<<<

 Extend Linear Regression to multiple dimensions. Many input and output sets with cross gain between each others, Just like a neural network layer.

>>>Correlation<<<

Pearson Correlation and my custom Linear Correlation metric for use in neural networks