A lot of people on Earth suffering from problems related to excess weight. And causes the appearance of extra pounds there are so many. However, artificial intelligence is connected to the solution of this problem, as the main factor of weight gain called not overeating or metabolic disorders, and errors in the development of urban neighborhoods.
According to the editors of the journal Science, a team of us researchers have taught AI to identify the areas most susceptible to the recruitment of extra pounds. And, based on data from more than 150,000 images from Google Maps, you AI tied, the percentage suffering from excess weight with the district planning and errors in urbanism (the science of development of urban systems).
Over the task worked 2 of the neural network. The first identified the key features of the areas, allowing to identify the distinctive features of districts, “coloring” them: green was designated trees, grey roads and red — houses and other buildings. The second neural network detected the connection between the configuration of the “painted” areas and indicators of obesity among people living in the area. Were analysed such US cities as Los Angeles, Memphis, San Antonio and Seattle.
For example, Memphis AI was able to identify that areas with a high degree of landscaping (top photo) associated with a low prevalence of obesity among residents. Conversely, where the amount of green space is less than (photo below)- the percentage of obesity higher.
Thus, according to the authors, the obtained data will allow not only to take better care of people’s health, but also allow business representatives to plan the infrastructure of their networks, placing the healthy food stores where required, and the designers of blocks in advance to think about the planting.
How do you feel about such studies? Tell us about it in our telegram chat.
AI decided that obesity is associated with… a wrong urban development
Vladimir Kuznetsov