No. Activities Milestones Deliverables Team members Duration of activity (from-to, in months)
1. Test multiple machine learning methods across a set of data to model energy consumption, CO2 emissions and costs Incentive models for predicting and classifying energy consumption, costs, and recovery periods across the entire set of public sector building objects (without clusters) Marijana Zekić-Sušac,

Josip Mesarić,

Dario Šebalj,

Adela Has,

Hrvoje Krstić,

Domagoj Sajter,

PhD student

m 13-15
2. For each identified cluster to make more precise prediction models for public buildings consumption and analysis of the input variables’ impact on energy consumption and CO2 emissions Predictive models for analyzing the influence of variables on energy management of public buildings by groups (clusters) Models for forecasting energy consumption of public buildings by groups (clusters) Marijana Zekić-Sušac, Rudolf Scitovski

Josip Mesarić,

Davor Dujak,

Dario Šebalj,

Adela Has,

Hrvoje Krstić,

Saša Mitrović,

PhD student

m 16-18
3. For each identified cluster to make more accurate estimates of public building costs and impact analysis of input variables on costs  Predictive models for analyzing the influence of variables on the management of energy costs of public buildings by groups (clusters) Models for Predicting Costs of Public Buildings by Groups (Clusters) Marijana Zekić-Sušac

Domagoj Sajter

Dario Šebalj

Adela Has

Hrvoje Krstić

m 19-21
4. Using statistical methods, compare the accuracy of machine learning methods in modeling energy consumption and costs The obtained comparison of the accuracy of the model with the tests of the significance of differences in the result and the like. Marijana Zekić-Sušac

Adela Has

Rudolf Scitovski

PhD student

m 22-24
5. Determine which methods are more successful in predicting spending for individual groups and suggest the integration of multiple methods  

Estimated Model Success and Comparison of Model Modeling Accuracy.

A choice of prediction and classification of energy consumption and costs has been made.

The most successful models for predicting energy consumption are chosen Marijana Zekić-Sušac

Josip Mesarić

Domagoj Sajter

Dario Šebalj

Adela Has

Hrvoje Krstić

Saša Mitrović

Rudolf Scitovski

PhD student

m 25-27
6. Develop a methodological framework for selecting appropriate methods and models Concept of the methodological framework Marijana Zekić-Sušac,

Josip Mesarić,

Dario Šebalj,

Adela Has,

Hrvoje Krstić,

Saša Mitrović,

Rudolf Scitovski,

Domagoj Sajter,

Davor Dujak

Zlatko Tonković

m 28-30
7. Formulate the methodological framework in the guidelines for the design of a forecasting model of energy consumption and costs  Publication of a methodological framework with guidelines in the form of algorithms for selecting methods in modeling The methodological framework algorithm published on the web site of the project Marijana Zekić-Sušac,

Josip Mesarić,

Dario Šebalj,

Adela Has,

Hrvoje Krstić,

Saša Mitrović,

Rudolf Scitovski,

Domagoj Sajter,

Davor Dujak

Zlatko Tonković

m 31-33
8. Prepare and maintain workshop “MERIDA Workshop 1” for users of consumption and cost forecasting models in Zagreb Workshop 1 – use of model for forecasting consumption and analysis of the influence of variables on energy consumption Workshop for employees of the Environmental Protection and Energy Efficiency Fund, the Agency for Legal and Real Estate Intermediation and Other Potential Beneficiaries of the Predicting Consumption Model and the Energy Consumption Alteration Analysis Marijana Zekić-Sušac

Hrvoje Krstić

Adela Has

Dario Šebalj, PhD student

m 34-36
9. Obtain literature (printed or e-books) from the field of machine learning, energy efficiency and financial modeling Acquired literature (printed or e-books) to study the methods and models All members of the team m 14-15
10. Training of team members at workshops in the field of cluster analysis – 2018 and 2019: Workshop at Leibniz Centrum Fuer Informatik, Dagstuhl, Germany Report on acquired new knowledge from workshops in the field of cluster analysis and machine learning Rudolf Scitovski Max. 8 days between m 13-24

 

Max. 8 days between m 25-36

11. Training of team members at workshops in the field of machine learning and analytical modeling Report on acquired new knowledge from workshops in the area of machine learning and analytical modeling Adela Has Max. 8 days between m 13-24

 

Max. 8 days between m 25-36

12. Training of a team member at a workshop on energy efficiency in the EU Report on acquired new knowledge from workshops in the field of energy efficiency of buildings Hrvoje Krstić Max. 8 days between m 25-36
13. Training a team member at a workshop in the field of machine learning and business analytics Report on acquired new knowledge from workshops in the area of machine learning and business analytics Marijana Zekić-Sušac Max. 8 days between m 25-36
14. Training of team members at a workshop in the field of machine learning and data analytics Report on acquired new knowledge from workshops in the area of machine learning and business analytics PhD student Max. 8 days between m 13-24

 

 

Max. 8 days between m 25-36

 

Max. 8 days between m 36-48

15. Creating and printing posters for a project that will be exhibited at conferences and workshops where project participants exhibit Poster (poster) about the project PhD student m 9-12

 

m  21-24

 

m 45-48

16. Writing and submitting papers on reading and then publishing in scientific journals – works from phase 2 of the project, which belongs to O2  Written scientific papers as a result of the project Scientific papers sent for publication in journals:

1 articles in the field of machine learning model comparison in modeling energy efficiency including cluster analysis (Marijana Zekić-Sušac, Rudolf Scitovski)
1 articles from the comparison of the classification methodology in energy efficiency modeling (Marijana Zekić-Sušac, Adela Has, Hrvoje Krstić)

Marijana Zekić-Sušac, Rudolf Scitovski,

Adela Has

 

 

 

 

Marijana Zekić-Sušac,

Adela Has, Hrvoje Krstić

Sending: m 13-24

 

Expected release time: m 24-36

17. Presentation of scientific papers at conferences from phase 2 of the project related to the O2 goal  Scientific papers exhibited at conferences Presented and published scientific papers in conference proceedings according to the plan:

3 works presented by Hrvoje Krstić at 3 conferences in the area of energy efficiency of buildings (1 work 2018, 1 work 2019 and 1 work 2020)
3 works by Adela Has at conferences in the area of machine learning and business analytics (1 paper 2018, 1 paper 2019 and 1 paper work 2020)
2 works by Marijana Zekić-Sušac in conferences in the area of machine learning and business analytics (1 work 2018, 1 paper 2019)
1 paper presented by Rudolf Scitovski at a conference on cluster analysis (2020)
2 papers present a PhD student at conferences in the area of machine learning and business analytics (1 work 2019 and 1 work in 2020)

 

 

 

 

Hrvoje Krstić

 

 

 

 

 

 

Adela Has

 

 

 

 

 

Marijana Zekić-Sušac

 

 

 

 

Rudolf Scitovski

 

 

 

PhD student

 

 

Other associates participate as co-authors depending on the contribution of the work

m 13-48
18. Defining topics of the doctoral dissertation from the project area  Defined topics for two doctoral dissertations from the project Defined and applied topics for two doctoral dissertations:

1 dissertation on the topic of accuracy analysis of machine learning methods in predicting energy efficiency of buildings
1 dissertation on public sector energy expenditure management

Adela Has

and

PhD student

m 13-24
19. Work on dissertations and complete dissertations from the project area  Two doctoral dissertations from the project were written Completed (written and submitted) at least 2 dissertations from the project area Adela Has,

PhD student

m 36-48