Sergei Markochev

In this competition, participants determine the aircraft positions based on time of arrival and signal strength measurements reported by many low-cost crowdsourced sensors. Only some receivers provide GPS-synchronized timestamps, while others experience strong clock drifts or provide fully broken timestamps.

The competition was organized by the Swiss Cyber-Defence Campus of Armasuisse Science and Technology. The data was collected by the OpenSky Network, a large-scale ADS-B sensor network for research.

• https://www.aicrowd.com/challenges/cyd-campus-aircraft-localization-competition/leaderboards

A data analysis of customer purchase history and prediction on their total spending in the future using Bayesian modelling and Monte Carlo simulation.

Notebook:
• Prediction of customer spending.ipynb

A Kaggle indoor competition aimed at predicting hotel sales for the first 10 days for a subset of new Expedia hotels (for which Expedia has no historical data).

I was ranked #1 among 19 teams proposing a combination of machine learning models.

Ticket prices change based on demand and time, and there can be a significant difference in price. In these two notebooks, I investigated the possibility of developing a pricing monitoring system for Spanish high-speed trains using data from Kaggle datasets.

Notebooks:
• Rail_ticket_price_prediction_IDE.ipynb
• Rail_ticket_price_prediction_modelling.ipynb

Drifting icebergs present threats to navigation and activities in areas such as offshore of the East Coast of Canada. In this competition, I was challenged to build an algorithm that automatically identifies if a remotely sensed target is a ship or iceberg.

Notebooks:
• Statoil_Kaggle_competition_main.ipynb
• Statoil_Kaggle_competition_google_colab_notebook.ipynb
• Statoil_Kaggle_competition_DL_comparison.ipynb