 |
About: currently, a full-time telecom/networks design engineer and data science coordinator at Dar.com and a part-time masters student in Data Science at the University of Nicosia. I sometimes do some engineering/data science tinkering at sodioum.com, independent research which I publish on arxiv.org and write in various blogs on klasssik.com.
Research:
- The Evolution of Decentralized ICT Networks: https://arxiv.org/abs/1909.04434
- Optionality and Convexity in ICT Networks: https://arxiv.org/abs/1909.13463
- Comparison Study for Multi-vendor Versus Single-vendor for Enterprise Computer Networks: https://arxiv.org/abs/2002.10777
Data Science Projects in BIM:
I'm always in search of doing tasks more efficiently (i.e. faster and better). Parametric and generative codes I developed using Python/Dynamo for BIM perform the following:
- Build a database of previous projects using SQL accompanied by a machine learning model which uses the decision tree algorithm on top of the database to design future projects. (read more: Role Models for BIM Models)
- Zoning and wiring elements using ML processes like the Markov Chain, k-means clustering and the traveling salesman algorithm. (Read more: traveling salesman in BIM, k-means clustering in BIM, Markov Chain in BIM)
- Find various MEP/AR elements and integrate the required ELV design and track future changes in the model automatically (read more: Data Design for CCTV Cameras)
Data Science Projects in Operational Buildings:
- Build ML algorithms to track the performance patterns of all systems to help provide future predictions of the systems targets and track deviations from these targets.
- Build an artificial neural network to detect faulty sensors readings
- Build a fault detection model for the PV system, AHU and elevators using machine learning algorithms and provide recommendations for performance optimization.
Outside working hours projects:
This is where I explore different topics of personal interest: