Flying Python at 40,000 feet
summary
Real black‑box multivariate time series get cleaned, aligned from HDF quick‑access dumps and segmented into taxi/takeoff/climb/cruise/descent; the concrete goal is end-of-takeoff detection using a Bayesian two‑step method that fits per‑flight posteriors then aggregates them for prediction. The pipeline uses R for exploration, Jupyter/Python for prototyping and production Python with numpy scipy pandas pymc (MCMC) to handle missing channels; repository FlightDataAnalyzer and 4,000,000 flights back the models and PyMC MCMC sampling provides the posterior samples used for per‑second probabilistic scoring.
AI-written summary · about
chapters (8)
- 0:00Flight data monitoring intro
- 2:56Flight data monitoring meaning
- 7:16Becoming flight data analyst
- 13:41Engine health monitoring challenges
- 15:27Takeoff end detection with Python
- 31:37Predicting takeoff events with MCMC
- 36:10Data confidentiality handling
- 39:16Future of data-driven aviation safety
related talks
18mSit back, relax, and enjoy the talk: a journey through FlightRadar24 with Python
42mMaking Sense Out of Flight Test Data with Python
43mRemove before flight: Analysing flight safety data.
32mBayesian Deep Learning and Flight Delay Prediction
42mRobust Automated Forecasting in Python and R
35mTouchdown Localisation with aircraft flight data
02h 00mData Science Driven Business Insights with Python and PySp
41mFootball's data analytics in python