Table 1.

Algorithm:

Step1: Initialize tde central node and site nodes.

Step2: Collect tde initial summary (number of samples) of tde site nodes in tde central node.

Step3: Calculate tde β coefficient for each site using tde decentralized regression approach available in COINSTAC.

Step4: For site node i = 1, 2, 3 …. N do

1. calculate tde local mean across tde features using tde local β coefficients

2. calculate tde local variance across tde feature using tde local β coefficients

3. send tde local mean and variance to tde aggregator node.

4. end for loop.

Step5: compute grand mean and grand variance and update each site node.

Step6: For site node i = 1, 2, 3 …. N do

1. standardized tde data w.r.t tde grand mean and grand variance.

2. estimate tde site parameters ${\gamma }_{i,v}{}^{*}and{\delta }_{i,v}^{2^{\ast }}$.

3. Adjust tde data accordingly.

4. Save tde adjusted data.

5. end for loop.