Short description
Summarize this content to 100 words NITU MISIS researchers have developed a program that predicts flight delays with high accuracy. On average, his forecasts differ from the true values by no more than 12 minutes. The app takes into account factors such as weather conditions, occupancy, type and age of air traffic, and aircraft maintenance issues. “Methods of machine learning perfectly cope with tasks in the field of air transportation. Determining the most important and informative features is a key stage in the development of an effective delay forecasting model,” shared the developer of the program, an employee of the Institute of Computer Sciences of NITU MRS Vyacheslav Pachkov.The program works on the basis of the multilayer perceptron model. She had been trained on a million flight records over the past year. MLP is capable of modeling more complex functions and dependencies between input and output data. It uses nine input features. These include the time between arrival and departure from the departure airport, expected arrival time at the destination airport, flight distance, departure airport, arrival airport, type of aircraft, temperature, probability of precipitation, season.Pachkiv will develop the application, increasing the accuracy of the model and optimizing the neural network to speed up work on weak mobile devices. He plans to release a full version for download.Meanwhile, Google has begun working with American Airlines and Bill Gates’ climate investment fund Breakthrough Energy to develop greener flight routes and reduce the impact of inversions left by planes.
NITU MISIS presented an application for tracking flight delays
NITU MISIS researchers have developed a program that predicts flight delays with high accuracy. On average, his forecasts differ from the true values by no more than 12 minutes.
The app takes into account factors such as weather conditions, occupancy, type and age of air traffic, and aircraft maintenance issues.
“Methods of machine learning perfectly cope with tasks in the field of air transportation. Determining the most important and informative features is a key stage in the development of an effective delay forecasting model,” shared the developer of the program, an employee of the Institute of Computer Sciences of NITU MRS Vyacheslav Pachkov.
The program works on the basis of the multilayer perceptron model. She had been trained on a million flight records over the past year. MLP is capable of modeling more complex functions and dependencies between input and output data. It uses nine input features. These include the time between arrival and departure from the departure airport, expected arrival time at the destination airport, flight distance, departure airport, arrival airport, type of aircraft, temperature, probability of precipitation, season.
Pachkiv will develop the application, increasing the accuracy of the model and optimizing the neural network to speed up work on weak mobile devices. He plans to release a full version for download.
Meanwhile, Google has begun working with American Airlines and Bill Gates’ climate investment fund Breakthrough Energy to develop greener flight routes and reduce the impact of inversions left by planes.
