University of Nevada, Reno

About: University of Nevada, Reno is a education organization based out in Reno, Nevada, United States. It is known for research contribution in the topics: Population & Poison control. The organization has 13561 authors who have published 28217 publications receiving 882002 citations. The organization is also known as: University of Nevada & Nevada State University.

A two-dimensional polymer prepared by organic synthesis.

Abstract: Synthetic polymers are widely used materials, as attested by a production of more than 200 millions of tons per year, and are typically composed of linear repeat units. They may also be branched or irregularly crosslinked. Here, we introduce a two-dimensional polymer with internal periodicity composed of areal repeat units. This is an extension of Staudinger's polymerization concept (to form macromolecules by covalently linking repeat units together), but in two dimensions. A well-known example of such a two-dimensional polymer is graphene, but its thermolytic synthesis precludes molecular design on demand. Here, we have rationally synthesized an ordered, non-equilibrium two-dimensional polymer far beyond molecular dimensions. The procedure includes the crystallization of a specifically designed photoreactive monomer into a layered structure, a photo-polymerization step within the crystal and a solvent-induced delamination step that isolates individual two-dimensional polymers as free-standing, monolayered molecular sheets.

Mentalism, behavior-behavior relations, and a behavior-analytic view of the purposes of science

Abstract: In a behavioral view, the purposes of science are primarily prediction and control. To the extent that a scientist embraces both of these as a unified and generally applicable criterion for science, certain philosophical and theoretical practices are counterproductive, including mentalism in both its metaphysical and metatheoretical forms. It is possible and often worthwhile to recast some mentalistic talk into an issue of behavior-behavior relations. When behavior-behavior relations are approached non-mechanistically, however, analysis cannot stop at the level of the relations themselves. Several analytic concepts common in the behavioral community share some of the dangers of mentalism if not employed properly, including such concepts as self-reinforcement, response-produced stimulation, and self-rules.

An insect-specific P450 oxidative decarbonylase for cuticular hydrocarbon biosynthesis

Abstract: Insects use hydrocarbons as cuticular waterproofing agents and as contact pheromones. Although their biosynthesis from fatty acyl precursors is well established, the last step of hydrocarbon biosynthesis from long-chain fatty aldehydes has remained mysterious. We show here that insects use a P450 enzyme of the CYP4G family to oxidatively produce hydrocarbons from aldehydes. Oenocyte-directed RNAi knock-down of Drosophila CYP4G1 or NADPH-cytochrome P450 reductase results in flies deficient in cuticular hydrocarbons, highly susceptible to desiccation, and with reduced viability upon adult emergence. The heterologously expressed enzyme converts C18-trideuterated octadecanal to C17-trideuterated heptadecane, showing that the insect enzyme is an oxidative decarbonylase that catalyzes the cleavage of long-chain aldehydes to hydrocarbons with the release of carbon dioxide. This process is unlike cyanobacteria that use a nonheme diiron decarbonylase to make alkanes from aldehydes with the release of formate. The unique and highly conserved insect CYP4G enzymes are a key evolutionary innovation that allowed their colonization of land.

Using Trusted Data to Train Deep Networks on Labels Corrupted by Severe Noise

Abstract: The growing importance of massive datasets with the advent of deep learning makes robustness to label noise a critical property for classifiers to have. Sources of label noise include automatic labeling for large datasets, non-expert labeling, and label corruption by data poisoning adversaries. In the latter case, corruptions may be arbitrarily bad, even so bad that a classifier predicts the wrong labels with high confidence. To protect against such sources of noise, we leverage the fact that a small set of clean labels is often easy to procure. We demonstrate that robustness to label noise up to severe strengths can be achieved by using a set of trusted data with clean labels, and propose a loss correction that utilizes trusted examples in a data-efficient manner to mitigate the effects of label noise on deep neural network classifiers. Across vision and natural language processing tasks, we experiment with various label noises at several strengths, and show that our method significantly outperforms existing methods.