Assessing the Free Land Programs for Reversing Rural Depopulation

A number of small towns in the Great Plains have recently started to offer free land and other incentives to entice new residents in the hope of reversing persistent depopulation. Based on in-depth interviews, this study assesses the initial performance of the free land programs in six small towns in central Kansas and analyzes the factors that have affected the migration decisions of the new residents. The initial results of these programs have been impressive. Not only have they attracted multiple new residents and increased enrollments in local schools, but they have also elevated long-time residents' pride in their community and created a positive synergy. The new residents' migration decisions were influenced by a number of push and pull factors. The free land and other incentives are not enough to trigger migration, but they have effectively changed some migrants' destination choice to a small town in central Kansas. Without the free land, most new residents, particularly those from out of state, would not have moved there. Contrary to our expectations, the relative locations of small towns with respect to larger cities do not appear to have affected new residents' destination choice.

Bandit-Based Algorithms for Budgeted Learning

We explore the problem of budgeted machine learning, in which the learning algorithm has free access to the training examples’ labels but has to pay for each attribute that is specified. This learning model is appropriate in many areas, including medical applications. We present new algorithms for choosing which attributes to purchase of which examples in the budgeted learning model based on algorithms for the multi-armed bandit problem. All of our approaches outperformed the current state of the art. Furthermore, we present a new means for selecting an example to purchase after the attribute is selected, instead of selecting an example uniformly at random, which is typically done. Our new example selection method improved performance of all the algorithms we tested, both ours and those in the literature.