Dosage individualization of erythropoietin using a profile-dependent support vector regression

The external administration of recombinant human erythropoietin is the chosen treatment for those patients with secondary anemia due to chronic renal failure in periodic hemodialysis. The objective of this paper is to carry out an individualized prediction of the EPO dosage to be administered to those patients. The high cost of this medication, its side-effects and the phenomenon of potential resistance which some individuals suffer all justify the need for a model which is capable of optimizing dosage individualization. A group of 110 patients and several patient factors were used to develop the models. The support vector regressor (SVR) is benchmarked with the classical multilayer perceptron (MLP) and the Autoregressive Conditional Heteroskedasticity (ARCH) model. We introduce a priori knowledge by relaxing or tightening the epsilon-insensitive region and the penalization parameter depending on the time period of the patients' follow-up. The so-called profile-dependent SVR (PD-SVR) improves results of the standard SVR method and the MLP. We perform sensitivity analysis on the MLP and inspect the distribution of the support vectors in the input and feature spaces in order to gain knowledge about the problem.     

Daily solar radiation from NASA-POWER product: assessing its accuracy considering atmospheric transparency

ABSTRACT

Satellite remote sensing in estimating solar energy budget components at the top of the atmosphere (TOA) level and at the terrestrial level plays a very important role in various types of applications. Solar radiation data are especially problematic because of a quite generalized lack of sufficient data in quantity and quality. Satellite images allow solving the problem of continuity or lack of solar radiation data. The objective of this work was to fit daily solar radiation from NASA-POWER (National Aeronautics and Space Administration – Prediction Of Worldwide Energy Resources), considering different intervals of atmospheric transparency index. The accuracy was assessed from the analysis of voluminous data-sets registered by meteorological ground stations, 31 in number, located in whole Spain, during the period from 2000 to 2017. Clearness index (K_T) was calculated to define nine classes of cloud cover conditions. The study reveals that the degree of correlation between the satellite data and observatory data depends upon atmospheric conditions and the correlation accuracy improves for higher values of K_T. The coefficients of determination (R²), considering all K_T values, were between 0.85 and 0.96; particularly for clear days R² = 0.96 and root-mean-square error equal to 1.78 MJ m^?2 d^?1 were obtained. Geographically, the better statistic values were located in the central region of the country. NASA-POWER shows potential to estimate solar radiation and that it is an important information resource for different applications.     

Estimating soybean ground cover from satellite images using neural-networks models

The ground cover is a necessary parameter for agronomic and environmental applications. In Argentina, soybean (Glycine max (L.) Merill) is the most important crop; therefore it is necessary to determine its amount and configuration. In this work, neural-network (NN) models were developed to calculate soybean percentage ground cover (fractional vegetation cover, fCover) and to compare the accuracy of the estimate from Moderate-Resolution Imaging Spectroradiometer (MODIS) and Landsat satellites data. The NN design included spectral values of the red and near-infrared (NIR) bands as input variables and one neuron output, which expressed the estimated coverage. Data of fCover were acquired throughout the growing season in the central plains of Córdoba (Argentina); they were used for training and validating the networks. The results show that the NNs are an appropriate methodology for estimating the temporal evolution of soybean coverage fraction from MODIS and Landsat images, with coefficients of determination (R ²) equal to 0.90 and 0.91, respectively.     

Development of regression equations for predicting energy and hygrothermal performance of buildings

Regression equations can be used for predicting indoor air temperature, relative humidity and energy consumption in an easier and more rapid way than building energy simulation tools. The independent variables, that is, the input data, are heating, ventilation and air conditioning (HVAC) power, outdoor temperature, relative humidity and total solar radiation. The present methodology for obtaining the regression equations is based on defining a couple of linear Multiple-Input/Single-Output (MISO) models, since two main outputs are involved, that is, indoor temperature and relative humidity. The methodology has been tested for the low- and high-thermal mass cases of the BESTest model (cases 600 and 900) and the output data is generated by using a building hygrothermal simulation tool. Validation procedures have shown very good agreement between the regression equations and the simulation tool for both winter and summer periods.     

Threshold-based identification of wireless SAW RFID-tags with pulse position encoding

In recent decades, there have been designed and developed passive surface acoustic wave radio frequency identification tags with time position encoding in groups of slots. For such tags identified when the received pulse-burst exceeds a threshold, we specify the generic identification (ID) error probability and optimize the threshold employing the Marcum Q-function of first order. As examples, we estimate the ID errors for the one-group 28-slot tag and Barker encoding. It is shown that, under the ideal conditions of equal SNRs (14.8 dB) in On-pulses and zeroth in Off-pulses, the ID error probability lesser 1% (one slip per 100 readings) can be guarantied with single reading for any reasonable number of slots. It becomes lesser 0.1% (1 slip per 1000 readings) with seven or lesser slots.     

Effect of lactobacilli inoculation on cassava (Manihot esculenta) silage: Fermentation pattern and kinetic analysis

The major effect of Lactobacillus inoculation on laboratory cassava (Manihot esculenta Crantz) silage was a change from a heterofermentative pattern observed in natural silage to a homofermentation. Small amounts of starter culture (1% v/w) were required to produce a high level of lactic acid (> 28 g kg^?1 DM) and to reach a pH of at least 4. The Gompertz model was used to evaluate the effect of inoculation level on the lactic formation based on kinetic criteria. Also an empirical pH lactic acid correlation was proposed to monitor the progress of ensiling, based solely on pH measurements. The simulation model may be used to improve guidelines for silo safety and to evaluate the effect of lactobacilli inoculants.     

Atypical antipsychotic medications: Pharmacological profiles and psychological implications.

Advances in psychopharmacology and the development of new antipsychotic medications may represent increased opportunities for psychologists to provide expanded psychosocial services to patients with schizophrenia. The new agents, referred to as the atypical antipsychotics, are as efficacious as the older conventional antipsychotics but demonstrate a more favorable side effect profile. Preliminary data suggest improvements in cognitive deficits and negative symptoms typically associated with schizophrenia, which may enhance long-term outcome. Psychologists may be increasingly called on to provide psychosocial services to this population, thereby requiring that they have a sound working knowledge of the pharmacological and psychological properties of these agents. (PsycINFO Database Record (c) 2011 APA, all rights reserved)     

On the coordinator's rule for Fast Paxos

Fast Paxos is an algorithm for consensus that works by a succession of rounds, where each round tries to decide a value v that is consistent with all past rounds. Rounds are started by a coordinator process and consistency is guaranteed by the rule used by this process for the selection of v and by the properties of process sets called quorums. We show a simplified version of this rule for the specific case where the quorums are defined by the cardinality of these process sets. This rule is of special interest for implementors of the algorithm.