README

The collection in this folder is described below:

iris.dat     this is the original data from UCI
irisN.dat    the above file was normalised so that the input attributes are all normalised by dividing throughout by the max attribute found

MushroomN.dat   same normalisation as above, division by greatest attribute. The character based data set was modified into a numeric one by giving each attribute a numeric value, rather than a letter value, so a, b, c would become 0, 1, 2 and so on.

Faults.dat   The original UCI data
FaultsN.dat  Normalisation by overall greatest attribute value
FaultsC.dat  Normalisation by greatest attribute with column
FaultsZMC.dat Normalisation by column and a zero mean for all input attributes

Blood1.dat    the original data, but with the output modified to give two output categories (representing a true one and a false one)
BloodN.dat    the above data but normalised for greates overall attribute
BloodC.dat   Normalised on a per attribute (Column) basis
BloodZMC.dat Normalised by column and has a zero mean

The following small programs are included so that anyone who is interested in python can check my work
The appropriate filenames and Input and Output sizes must be set within the programgedit for the data samples of interest
NormaliseN.py	just an overall input normalisatio
NormaliseC.py	normalisation on a per attribute basis
NormaliseZMC.py	normalisation per attribute and with zero mean