How can you learn to apply machine learning methods, set a problem, choose a model, find data for training it and simplify the operation of airports in a couple of months, having found a connection between stock indices and the daily number of passengers? Easier than it sounds.
Our team has been developing applications for more than ten years that control the operation of the largest airports: Frankfurt, Dublin, Manila, Jakarta, Miami, Beijing. Airports use applications for optimal resource management, organization of work and control over the flow of airport information, and coordination of flight schedules.
, . , 2019 70 , . , 190 , , . . . : , , . , , , . , . , . .
: « , – , : , , ». , . , , . , .
, . :
. . , .
. , .
. , , , ( ). , , .
. , , . DAX, .
DAX
Deutscher Aktienindex – German stock index – . .
DAX, Yahoo Finance. 2018 2019 , kaggle. , . , .
. , . DAX – . DAX DAX . 5 : 15 19 .
. MAE, ( / ). , DAX , . Linear Regressor, Random Forest, Gradient Boosting, SGDRegressor. GridSearchCV.
MAE
Mean Average Error
DAX . . () (). MAE . . MAE, score ( DAX + ) .
, DAX 15 . , , MAE. Gradient Boosting 297 . , .
pickle, REST API Docker-. , 15 , DAX, , . . .
. - , . , , .