As a result of research project MERIDA in 2018, several clustering and prediction models were developed, namely:

  • the clusters of public buildings in Croatia using modified DBSCAN algorithm (published in Croatian Operational Research Review (CRORR) journal, 2018), while the algorithm searching for an optimal k-partition was accepted to be published in Journal of Global Optimization
  • the models for predicting energy efficiency of public buildings based on artificial neural network methodology (published in Croatian Review of Economic, Business and Social Statistics (CREBSS) journal),
  • the model for predicting the electricity consumption based on tree-partitioning methods, among which the random forest was the most accurate one (published in CECIIS conference, Varazdin, 2018)
  • the model for predicting the natural gas consumption based on tree-partitioning methods (published in EBS conference on Resilient Society, Zagreb 2018)
  • the model for estimating financial effects of construction measures based on panel data analysis (published at EBS conference, Paris, France)
  • the algorithm for cleansing data on constructional characteristics of public buildings (presented at the HPSM/OPTI 2018, Ljubljana, Slovenia conference and published in the WIT Transactions on The Built Environment journal, 2018)

Also, regarding the modeling of natural gas supply chain, the following results were made:

  • the mapping of natural gas distribution in Croatia was conducted (and published at the  ICIL conference, Bar Sheeba, Israel, 2018)
  • the simulation model of natural gas supply was created (presented and published at the IBIMA conference, Seville, Span, 2018)
  • the Bullwhip Effects in Natural Gas Supply Chain were estimated (presented at the SENCERE conference, Dubrovnik, 2018)

That’s not all! See the list of all publications  resulted from the research project MERIDA.