Abstract: The near infrared (NIR) spectra technique is an effective approach to predict chemical properties and it is typically applied in petrochemical, agricultural, medical, and environmental sectors. NIR spectra are usually of very high dimensions and contain huge amounts of information. Most of the information is irrelevant to the target problem and some is simply noise. Thus, it is not an easy task to discover the relationship between NIR spectra and the predictive variable. However, this kind of regression analysis is one of the main topics of machine learning. Thus machine learning techniques play a key role in NIR based analytical approaches. Pre-processing of NIR spectral data has become an integral part of chemometrics modeling. The objective of the pre-processing is to remove physical phenomena (noise) in the spectra in order to improve the regression or classification model. In this work, we propose to reduce the noise using extreme learning machines which have shown good predictive performances in regression applications as well as in large dataset classification tasks. For this, we use a novel algorithm called C-PL-ELM, which has an architecture in parallel based on a non-linear layer in parallel with another non-linear layer. Using the soft margin loss function concept, we incorporate two Lagrange multipliers with the objective of including the noise of spectral data. Six real-life dataset were analyzed to illustrate the performance of the developed models. The results for regression and classification problems confirm the advantages of using the proposed method in terms of root mean square error and accuracy.

Abstract: Gaps in protoplanetary disks have long been hailed as signposts of planet formation. However, a direct link between exoplanets and disks remains hard to identify. We present a large sample study of ALMA disk surveys of nearby star-forming regions to disentangle this connection. All disks are classified as either structured (transition, ring, extended) or nonstructured (compact) disks. Although low-resolution observations may not identify large-scale substructure, we assume that an extended disk must contain substructure from a dust evolution argument. A comparison across ages reveals that structured disks retain high dust masses up to at least 10 Myr, whereas the dust mass of compact, nonstructured disks decreases over time. This can be understood if the dust mass evolves primarily by radial drift, unless drift is prevented by pressure bumps. We identify a stellar mass dependence of the fraction of structured disks. We propose a scenario linking this dependence with that of giant exoplanet occurrence rates. We show that there are enough exoplanets to account for the observed disk structures if transitional disks are created by exoplanets more massive than Jupiter and ring disks by exoplanets more massive than Neptune, under the assumption that most of those planets eventually migrate inwards. On the other hand, the known anticorrelation between transiting super-Earths and stellar mass implies those planets must form in the disks without observed structure, consistent with formation through pebble accretion in drift-dominated disks. These findings support an evolutionary scenario where the early formation of giant planets determines the disk's dust evolution and its observational appearance.