Analysis of emergency sites and, based on Machine Learning, recommendations on the location of rescue services
About project
NeuroMap is a system that predicts the likelihood of accidents at any time and provides guidance on the deployment of security services for a quick response to the situation. The system is based on Machine Learning technology and is a product that receives information about incoming calls to service 112, displaying a heat map of possible incidents and the placement of security services machines, depending on the level of danger.
Technical solutions
System structure
We have worked out the optimal structure of the system:
1. The system was trained on the basis of calls to 112 for 1 year in Moscow
2. The system provides a forecast for every hour and every day of the week.
3. The ability to arrange the machine, that is, to take the place in advance for quick response in case of an emergency
Machine Learning
Based on data from the call to service 112 in Moscow for 1 year, a forecast of the probability of occurrence of an emergency at each hour and day of the week is implemented
Thanks to the receipt of information about new calls, training the system and making changes in the work
Based on the forecast, response services are placed with the greatest probability of occurrence of an emergency.
Presentation of information
The view is implemented as a web interface and presents a heat map of the occurrence of incidents in each time period.
There is a functional to arrange security services, on the basis of the forecast, the coordinates of the services appear
There is a function to introduce a new point for retraining the system.
Technology benefits
Training on real data
The system is trained on real data and is updated all the time, which improves the accuracy of information
Location recommendations
The system provides rescue services location endpoints on the map for the fastest update.
The system is constantly updated and change shoes, which allows you to stay current all the time.
The project team. Winners NeuroHive Hackathon 2018.
Get more product information
Contact the team
Watch a video demo
Product demo
+7 965 241 27 08