Learn More
This study investigates how parental employment affects child cognitive development. The results indicate that maternal labor supply during the first three years of the child's life has a small negative effect on the predicted verbal ability of 3 and 4 year olds and a larger detrimental impact on the reading and mathematics achievement of 5 and 6 year olds(More)
  • Robert W Fairlie, Bruce D Meyer, Joseph Altonji, Rebecca Blank, Thomas Dunn, Lori Kletzer +7 others
  • 1999
We examine trends in self-employment among white and black men from 1910 to 1990 using Census and CPS microdata. Self-employment rates fell over most of the century and then started to rise after 1970. For white men, we find that the decline was due to declining rates within industries, but was counterbalanced somewhat by a shift in employment towards high(More)
The views expressed by the authors of this paper are their own, and do not reflect the opinions of any organization with which they may be affiliated. Abstract This paper investigates the impact of financial incentive programs, which have become an increasingly common component of welfare programs. We review experimental evidence from several such programs.(More)
Note: The authors are collaborators on a study of inequality in early childhood education and care, funded by the Russell Sage Foundation as part of its Social Inequality program. We would like to thank Eric Wanner and the Foundation for their support and would also like to thank Jay Bainbridge, Se-Ook Jeong, Katherine Magnuson, and Sakiko Tanaka for their(More)
In recent years, approaches based on machine learning have achieved state-of-the-art performance on image restoration problems. Successful approaches include both generative models of natural images as well as discriminative training of deep neural networks. Discriminative training of feed forward architectures allows explicit control over the computational(More)
Cameras that can measure the depth of each pixel in addition to its color have become easily available and are used in many consumer products worldwide. Often the depth channel is captured at lower quality compared to the RGB channels and different algorithms have been proposed to improve the quality of the D channel given the RGB channels. Typically these(More)
Optical flow is typically estimated by minimizing a " data cost " and an optional regularizer. While there has been much work on different regularizers many modern algorithms still use a data cost that is not very different from the ones used over 30 years ago: a robust version of brightness constancy or gradient constancy. In this paper we leverage the(More)
The goal of subspace learning is to find a k-dimensional subspace of R d , such that the expected squared distance between instance vectors and the sub-space is as small as possible. In this paper we study the sample complexity of subspace learning in a partial information setting, in which the learner can only observe r ď d attributes from each instance(More)