2016 - 2018: Diploma Thesis
Customer: FIIT STU
Tasks:
Topic on request
- The aim of this thesis is to optimize energy usage at smart homes. Optimization of this kind is not a trivial problem and opens the number of various challenges.
- Because of the fact that it uses energy usage and energy generation data, which are not simple linear functions, it is needed to take a lot of restrictions and constraints into consideration.
- The aim is to maximize the usage of generated power to prevent energy leakage and to minimize energy usage from the smart grid, in order to save some money. We may use an energy accumulator, like a battery, for temporary storage of wasted energy from generation.
- Afterwards the energy management optimization at homes is plausible.
- After the good optimization algorithms are chosen and the optimization model is set, it is inevitable to test it on real world data to prove and evaluate the proposed concept.
- I learnt working with scientifi c library packages written in python and had to adopt several core concepts of machine learning algorithms.
- I was experimenting with data, confi guring ML algorithms and visualizing the results on daily bases.
Skills:
Python, Data science, PSO, L-BFGS-B
2017 - 2018: My Food Court Team project
Tasks:
- My food court is an application in which you may view your favorite lunch meals in your surroundings.
- It was a university project of 6 people taking one year long.
- The purpose was to learn communication in a team, resolving issues and discussing new features in an agile environment following SCRUM.
- We had to face technological challenges like harvesting lunch meals from the web, parsing and normalizing unstructured data, building recommendation engines and using machine learning for getting new pieces of information.
Skills:
Angular, Python, Data science, Neural network, Postgres, PHP