[DSC 4.0] Implementation of genetic algorithm in problem of processing GPS data - Marko Nikolic

During the first part of the talk, brief theory about genetic algorithm will be presented, as well as its implementation in process of data mining and optimization. The second part will introduce the problem of accurate processing of GPS data of the vehicles, which is important aspect for reconstructing the route and the position of vehicles on the street network. The third part will introduce and present novelty methodology proposed in paper, in which is used genetic algorithm, dynamic time wrapping and dynamic programming techniques (the main part of the talk). Beside this, this part will present useful code examples made in Python of some techniques used in genetic algorithm and research. Final part will present results of tests and findings of research. Model is tested on real data collected from the street network of Belgrade, which yielded good results with regard to accuracy (the most accurate model for sparse GPS data) and running time.

This problem may concern public who are interested in automation vehicles, navigation on maps and spatial analyses. Moreover, because genetic algorithm is usually used as method for optimization, one of the interesting part of this research is that this algorithm is used for processing the data instead.

This talk was presented by Mr. Marko Nikolic, Data Scientist at Bravo Systems, during Data Science Conference 4.0, as a part of ML & AI track.

You can find this talk presentation on the following link:

More info about Data Science Conference:

To watch more new videos regarding Data Science - click subscribe to our YouTube Channel.