Data Science Economy: Data Driven
3 keynotes / 16 talks / 32 speakers
Meet the Speaker!

17. – 18.05. 2018, @HUB385, Zagreb, Croatia
Co-Owner & Chief Technology Officer, Business Intelligence / Big Data @NEOS

Mario Mrljić

Mario Mrljić is one of the founders and co-owner of Neos Ltd and currently holding the position of CTO.
Over 15 years of experience based on number of domestic and international DW/BI projects providing consulting services related to different areas ranging from high-level system architecture design to technical implementation of individual modules.
Interested in following areas:
– Business Intelligence, Data Warehouse & Big Data Systems
– Profitability Analysis & Risk Management Solutions
– Financial Institutions & Telco Performance Measurement



Follow Mario




Sessions and Workshops

Mario will have one presentation on the topic of forecasting natural gas and this is great case study with the customer!


Forecasting natural gas consumption using predictive modelling

IProject has been realized with Gradska plinara Zagreb Ltd, leading gas distributor in Croatia. In short it can be described as step closer to understand the future. Prediction Module of natural gas consumption provides multiple benefits when it comes to management and planning of distribution network. It is a complex solution consisting of:
• Data extraction,
• Data Modelling (DWH like),
• Data Analysis,
• Predictive Modelling,
• Results Validation,
• Web base Application,
• Reporting Module.

Starting with data extraction, statistical and exploratory analysis alongside with predictive modeling in Oracle R best algorithms and parameters have been defined to support calculation module.

Scheduler functionalities and web application Analytics Insight provided monitoring of all automated processes and overview of predictions and validation results.

Reporting module that has been implemented allows overview of gas consumption, comparison of predicted versus realized, trends in consumption, etc. Project resulted with fully automated prediction on hourly level with deviation of max. 2-3%.

Read more