Desperate for a better life, be it relief from war and persecution or simply an escape from grinding poverty and lack of opportunity, huge numbers of people from the Middle East and Africa have sought refuge in Europe in recent years - around 1.8 million of them since 2014. Sometimes their dreams of sanctuary and welcome are realised; they make it to the continent, are granted leave to stay and, with help, begin the slow and difficult process of establishing a place for themselves. And sometimes it all goes terribly wrong; the dangers, setbacks and obstacles on the journey are simply too great to overcome and they are forced to give up or turn back, or the reception they get on arrival is so hostile and unforgiving that eventually they are deported or disenchantment drives them home. In the first of two consecutive episodes exploring these contrasting experiences, People & Power has been to the small affluent city of Detmold, in north west Germany, the European country which under the government of Chancellor Angela Merkel, has taken in more refugees and migrants than any other and which, through generous state-funded welfare provision and language and job training, has sought to make a success of integration. Although this process is by no means universally popular across Germany - anti-migrant sentiment found in other parts of Europe is increasingly being echoed here too - in Detmold, at least, where the openheartedness of the local population is making a crucial difference, it appears to be working.
As University of Washington economist Jacob Vigdor summed up in his research on recent immigrants, fears of a lack of assimilation in the United States are overblown. "Basic indicators … from naturalization to English ability, are if anything stronger now than they were" in the Ellis Island era. The law guaranteeing birthright citizenship is part of the reason. Far from ridiculous, it guarantees that immigrants and their children are woven tightly into the American fabric. Let's keep it in place, and the 14th Amendment intact.
This paper describes a method and system for integrating machine learning with planning and data visualization for the management of mobile sensors for Earth science investigations. Data mining identifies discrepancies between previous observations and predictions made by Earth science models. Locations of these discrepancies become interesting targets for future observations. Such targets become goals used by a flight planner to generate the observation activities. The cycle of observation, data analysis and planning is repeated continuously throughout a multi-week Earth science investigation.
After Donald Trump signed his first executive order to ban refugees and immigrants from a number of Muslim-majority nations in January, I tweeted this photo of my dad riding the New York City Subway in a righteous afro and bell-bottoms. The point was to mock the notion that Muslims aren't capable of assimilating in the West. This is a sentiment I hear more than any other about why people like me don't belong in America--more than "vetting" dog whistles, more than fantasies about sleeper cells. It's one that reportedly has prominent sway in the White House right now. According to the Los Angeles Times, two of Trump's top advisers, Steve Bannon and Stephen Miller, "have pushed an ominous view of refugee and immigration flows, telling other policymakers that if large numbers of Muslims are allowed to enter the U.S., parts of American cities will begin to replicate marginalized immigrant neighborhoods in France, Germany and Belgium."
We identify a strong equivalence between neural network based machine learning (ML) methods and the formulation of statistical data assimilation (DA), known to be a problem in statistical physics. DA, as used widely in physical and biological sciences, systematically transfers information in observations to a model of the processes producing the observations. The correspondence is that layer label in the ML setting is the analog of time in the data assimilation setting. Utilizing aspects of this equivalence we discuss how to establish the global minimum of the cost functions in the ML context, using a variational annealing method from DA. This provides a design method for optimal networks for ML applications and may serve as the basis for understanding the success of "deep learning". Results from an ML example are presented. When the layer label is taken to be continuous, the Euler-Lagrange equation for the ML optimization problem is an ordinary differential equation, and we see that the problem being solved is a two point boundary value problem. The use of continuous layers is denoted "deepest learning". The Hamiltonian version provides a direct rationale for back propagation as a solution method for the canonical momentum; however, it suggests other solution methods are to be preferred.