A New Theoretical Approach: A Model Construct for Fault Troubleshooting in Cloud Computing

Algorithm 1

A proposed algorithm for transforming numerical data to binning.

Input: Metrics_values , mean, m, r,    //mean[] is a list of summation columns by

n. num. index generated from step 1, calculated from

Eq. (10), and are predetermined statistically

in this experiment and , where, is

the value range [], is equal-sized bins

indexed from 0 to . // size of look-back

window index for row table for normalized

values table. //num of the number of

components metrics index for column

normalized values table. //Normalized is

the normalization values table. //Bin is the

a bin index data binning table. //Filtering-

dataset is function algorithm to remove and

process outliner and noises data.

Output: Data binning table.

() For  , ; i++ do

()  For  , <= num,   j++  do

()   Normalized = Metrices_values/mean

()   If  Normalized   then

()      = m

()   else

()     =TRANC(Normlized/(r/m))

()  End for

() End for

() Data_binning=Filtering-dataset(Bin)

() End Function return (Data_binning).