Centralization and decentralization of SPSS
2. Forced centralization or centralization, only the non standardization coefficient is different, the standardization system is the same
(provided by Nanxin)
The purpose of centralization is to unify the units, that is, to unify the dimensions, because the units of different variables are different, which will cause the errors of various statistics
first calculate the average value of variables
in this way, the work of centralizing variables is completed
analysis description statistics description, and then select "save standardized score as a variable" and confirm to get the processed standardized data, and then cluster, factor and regression analysis can be carried out
subtract the average value: first perform a description statistics to get the descriptive statistical results, with the average and standard deviation. Then use the compute command to create a new variable = original variable average
the Z-score is similar to the above result, except that the new variable is divided by the standard deviation to get a score
the question is your description: a variable has multiple items. What does that mean? I can't think of it.
2. The first step is for the independent variables to enter the regression equation; The second step is that independent variables and regulatory variables enter together; The third step is that the independent variable, regulatory variable and interaction item enter together
3. The adjustment variables were divided into high and low groups, and the regression analysis of independent variables and dependent variables was done. Then the influence coefficient of high and low groups of independent variables on dependent variables was compared, and the slope test was carried out
1. Efficiency is different: the efficiency of decentralized network is higher than that of centralized network, but conversely, the security of centralized network is higher. On the whole, centralized network is suitable for high security applications
2. Different characteristics: centralized network means that each time two client computers communicate with each other through the server. The decentralized network is the initial need to go through the server, later communication does not need the server
3. Different costs: the characteristics of collective maintenance of blockchain can rece a lot of costs. The centralized network involves the maintenance, confidentiality, operation and so on of many systems, which must require a lot of human, financial and material resources. These things do not exist for the decentralization of blockchain, so they can save a lot of costs
extended materials:
precautions:
in the aspect of blockchain governance, an interesting trend recently is the re emergence of online virtual currency holder voting as a multi-purpose decision-making mechanism. Sometimes the vote of the coin holder is used to decide who runs the super node of the network (e.g. dpos in EOS, Neo, lisk and other systems), sometimes the protocol parameters are voted (e.g. the upper limit of Ethereum fuel supply), and sometimes the vote and direct application of approval (e.g. tezos) protocol upgrade are also concted
in all these cases, voting is automatic, and the protocol itself contains all the logic needed to change the verifier set or update its own rules, and automatically responds to the voting results