No abstract

We apply instance-based machine learning in the form of a k-nearest neighbor algorithm to the task of estimating photometric redshifts for 55,746 objects spectroscopically classified as quasars in the Fifth Data Release of the Sloan Digital Sky Survey. We compare the results obtained to those from an empirical color-redshift relation (CZR). In contrast to previously published results using CZRs, we find that the instance-based photometric redshifts are assigned with no regions of catastrophic failure. Remaining outliers are simply scattered about the ideal relation, in a similar manner to the pattern seen in the optical for normal galaxies at redshifts z 1. The instance-based algorithm is trained on a representative sample of the data and pseudo-blind-tested on the remaining unseen data. The variance between the photometric and spectroscopic redshifts is σ 2 = 0.123 ± 0.002 (compared to σ 2 = 0.265 ± 0.006 for the CZR), and 54.9 ± 0.7%, 73.3 ± 0.6%, and 80.7 ± 0.3% of the objects are within ∆z < 0.1, 0.2, and 0.3 respectively. We also match our sample to the Second Data Release of the Galaxy Evolution Explorer legacy data and the resulting 7,642 objects show a further improvement, giving a variance of σ 2 = 0.054 ± 0.005, and 70.8 ± 1.2%, 85.8 ± 1.0%, and 90.8 ± 0.7% of objects within ∆z < 0.1, 0.2, and 0.3. We show that the improvement is indeed due to the extra information provided by GALEX, by training on the same dataset using purely SDSS photometry, which has a variance of σ 2 = 0.090 ± 0.007. Each set of results represents a realistic standard for application to further datasets for which the spectra are representative.

An outstanding challenge in biomedical sciences is to devise a palette of molecular probes that can enable simultaneous and quantitative imaging of tens to hundreds of species down to ultralow concentrations. Addressing this need using surface-enhanced Raman scattering-based probes is potentially possible. Here, we theorize a rational design and optimization strategy to obtain reproducible probes using nanospheres with alternating metal and reporter-filled dielectric layers. The isolation of reporter molecules from metal surfaces suppresses chemical enhancement, and consequently signal enhancements are determined by electromagnetic effects alone. This strategy synergistically couples interstitial surface plasmons and permits the use of almost any molecule as a reporter by eliminating the need for surface attachment. Genetic algorithms are employed to optimize the layer dimensions to provide controllable enhancements exceeding 11 orders of magnitude and of single molecule sensitivity for nonresonant and resonant reporters, respectively. The strategy also provides several other opportunities, including a facile route to tuning the response of these structures to be spectrally flat and localization of the enhancement within a specific volume inside or outside the probe. The spectrally uniform enhancement for multiple excitation wavelengths and for different shifts enables generalized probes, wheras enhancement tuning permits a large dynamic range by suppression of enhancements from outside the probe. Combined, these theoretical calculations open the door for a set of reproducible and robust probes with controlled sensitivity for molecular sensing over a concentration range of over 20 orders of magnitude.Raman spectroscopy | surface enhanced | chemical imaging | vibrational spectroscopy S urface-enhanced Raman scattering (1) (SERS)-based probes, consisting of nanostructured particles, are strongly emerging for biomedical applications. SERS-based probes (2, 3) are exceptionally attractive as they offer quantitative enhancement of signal with facile readout (4, 5), extensively multiplexed imaging (6), and ultrasensitive assays (7, 8)-but not all at the same time. The SERS effect is typically prominent in nanoscale metal-dielectric environments in which the signal of a proximal organic molecule can be rationally tailored (9) and enhanced to the extent that single molecules may be detected (10, 11). Hence, SERS probes typically contain nanoscale metallic structures and organic molecules (12) that act as a quantitative reporter for the presence of the probe. The signal of this reporter is greatly enhanced to transduce biochemical species of low concentration at the molecular (13), cellular (14), and tissue levels (15, 16) to measurable signals. The achieved enhancement depends on the reporters' molecular characteristics as well as nanoscale size, shape, geometry, local aggregation state, and surface characteristics of the metal. These parameters can potentially be controlled to tune the reporters' signal, especially to m...

No abstract