"SOLUTIONS" Tag Cloud Globe (innovation business projects ideas)


The analytical capacity of BayesiaLab software is optimized in terms of identification of

(i) points of best contribution and optimization of a selected variables

(ii) new parameters to be carefully taken into consideration and regularly monitored for improving programmes performances.

 The proposed analytical services address the reliability of modeling multi-level hierarchical systems. The models developed within the framework of Public Health Programmes  and Health System Strengthening  could be used to effectively represent the true structure of health systems components and more importantly, the interdependency among parameters and innovative analyses of successes or weak contribution causes.

Map of Analytic Modeling

Bayesian networks can cover the entire map of analytics. The figure shows what this means in practice for the researcher : BayesiaLab’s functions, represented as blue boxes, are positioned across this map, and they demonstrate the applicability of the Bayesian network paradigm to « Surveys » or « Public Health programmes ».


Chart 3

Tags on the x-axis furthermore indicate a conceptual progression from description to optimization. Target or impact indicators optimization can be easily studied within the framework of Public Health programmes.

The y-axis shows the source of the model specification: On the one end, we have Theory, sourced from Human Intelligence (such as the six WHO building blocks or the agreed relationship among indicators as deemed in a performance framework or log frame).

On the other end, we have Data as the source. Data is associated with Machine Learning and Artificial Intelligence. It is closely connected with the processes related to the first service offer : Analyze of surveys.



The managerial processes, leading to recommendations to decisions makers, are laid down because of the (causal) relationships among the parameters and contributing factors in relation to programme performances.

Assuming that the model, the parameters and their interactions and finally their probabilistic relationships with program performances are correctly validated in the previous step,  “computerized model” can yield a deep understanding of a public health programs indeed.

It is mainly a question of semantic or meaning and, the main challenge will be:

  1. to construct rich analytical narratives that avoid oversimplification
  2. to establish temporal and plausible relationships
  3. to recognise path dependency among parameters

Overall, to translate the results of probabilistic functions or formulas into common language


Last but not least, the model applied to health programme could represent an innovative knowledge framework and could provide a common reasoning language between stakeholders from different backgrounds, such as international donors or policy makers on one side and, research experts on the other side. As such, with all available knowledge united, properly communicated and put into a “reasonable” format, the model could be a useful tool for making decisions and shaping policies.

The simulation capacity of the software could also enable to appraise the consequences of actions, decision makers might take.