Predicting quality of an information system using system dynamics modelling and machine learning

An apparatus for performing predictive analysis on an external index number describing quality of an information system comprising one or more of information technology systems, communications systems and access networks and producing information quantifiable by a plurality of quality meters, the apparatus comprising:

• at least one processor, and
• at least one memory comprising a computer program code, wherein the processor, the memory, and the computer program code are configured to cause the apparatus to perform:
• maintaining, in the at least one memory, a dynamic causal loop diagram of the plurality of quality meters describing the information system, wherein the dynamic causal loop diagram of the plurality of quality meters comprises present information on values of the plurality of quality meters and interactions with their strengths between the plurality of quality meters;
• maintaining, in the at least one memory, history data comprising information on past and present values of the plurality of quality meters and the external index number and interactions with their strengths between the plurality of quality meters;
• selecting one or more quality meters from the plurality of quality meters of the dynamic causal loop diagram for predicting values of the external index number, wherein the selecting is based at least on the dynamic causal loop diagram;
• processing values of the one or more quality meters and the external index number comprised in the history data using one or more statistical analysis methods to determine dependencies between the plurality of quality meters and the external index number;
• establishing, based on the processing, a prediction model for predicting external index number values based on one or more sets of values of the one or more quality meters;
• in response to receiving training data comprising at least one value for the external index number and one corresponding set of values for the one or more quality meters, training the prediction model using one or more machine learning tools based on received training data;
• calculating, using the prediction model, a present predicted external index number value based on present values of the one or more quality meters;
• and causing displaying the present predicted external index number to one or more users via one or more displays of one or more user equipment for managing quality of the information system.

A non-transitory computer-readable medium comprising program code instructions for a data processing apparatus of performing predictive analysis on an external index number describing quality of an information system comprising one or more of information technology systems, communications systems and access networks and producing information quantifiable by a plurality of quality meters, wherein execution of the program code instructions causes the data processing apparatus to perform the acts of:

• maintaining, in a database, a dynamic causal loop diagram of the plurality of quality meters describing the information system, wherein the dynamic causal loop diagram of the plurality of quality meters comprises present information on values of the plurality of quality meters and interactions with their strengths between the plurality of quality meters;
• maintaining, in the database, history data comprising information on past and present values of the plurality of quality meters and the external index number and interactions with their strengths between the plurality of quality meters;
• selecting one or more quality meters from the plurality of quality meters of the dynamic causal loop diagram for predicting values of the external index number, wherein the selecting is based at least on the dynamic causal loop diagram;
• processing values of the one or more quality meters and the external index number comprised in the history data using one or more statistical analysis methods to determine dependencies between the plurality of quality meters and the external index number;
• establishing, based on the processing, a prediction model for predicting external index number values based on one or more sets of values of the one or more quality meters;
• in response to receiving training data comprising at least one value for the external index number and one corresponding set of values for the one or more quality meters, training the prediction model using one or more machine learning tools based on received training data;
• calculating, using the prediction model, a present predicted external index number value based on present values of the one or more quality meters;
• and
• causing displaying the present predicted external index number to one or more users via one or more displays of one or more user equipment for managing quality of the information system.