Search

Seizure prediction using EEG has significant implications for the daily monitoring and treatment of epilepsy patients. However, the task is challenging due to the underlying spatiotemporal correlations and patient heterogeneity. Traditional methods often use large-scale models with independent components to capture the spatial and temporal features of EEG separately or explore shared patterns among patients with the help of pre-defined functional connectivity. In this paper, we propose a compact model, called the graph convolutional network based on adaptive functional connectivity (AFC-GCN), for seizure prediction. The model can adaptively infer evolution of functional connectivity in epilepsy patients during seizures through data-driven methods and synchronously analyze spatiotemporal response of functional connectivity in multiple topologies. On CHB-MIT datasets, the experimental results demonstrate that AFC-GCN achieves accurate and robust performance with low complexity. (AUC: 0.9820, accuracy: 0.9815, sensitivity: 0.9802, FPR: 0.0172). The proposed method has the potential to predict seizure during daily monitoring.

Background: The electrophysiology algorithm for localizing left or right origins of outflow tract ventricular arrhythmias (OT-VAs) with lead V 3 transition still needs further investigation in clinical practice., Hypothesis: Lead I R-wave amplitude is effective in distinguishing the left or right origin of OT-VAs with lead V 3 transition., Methods: We measured lead I R-wave amplitude in 82 OT-VA patients with lead V 3 transition and a positive complex in lead I who underwent successful catheter ablation from the right ventricular outflow tract (RVOT) and left ventricular outflow tract (LVOT). The optimal R-wave threshold was identified, compared with the V 2 S/V 3 R index, transitional zone (TZ) index, and V 2 transition ratio, and validated in a prospective cohort study., Results: Lead I R-wave amplitude for LVOT origins was significantly higher than that for RVOT origins (0.55 ± 0.13 vs. 0.32 ± 0.15 mV; p < .001). The area under the curve (AUC) for lead I R-wave amplitude as assessed by receiver operating characteristic (ROC) analysis was 0.926, with a cutoff value of ≥0.45 predicting LVOT origin with 92.9% sensitivity and 88.2% specificity, superior to the V 2 S/V 3 R index, TZ index, and V 2 transition ratio. VAs in the LVOT group mainly originated from the right coronary cusp (RCC) and left and right coronary cusp junction (L-RCC). In the prospective study, lead I R-wave amplitude identified the LVOT origin with 92.3% accuracy., Conclusion: Lead I R-wave amplitude provides a useful and simple criterion to identify RCC or L-RCC origin in OT-VAs with lead V 3 transition., (© 2020 The Authors. Clinical Cardiology published by Wiley Periodicals LLC.)

Biomarker selection has played an increasingly important part in modern medicine with advances of omics techniques. Kohonen self-organizing map is a well-established variable reduction algorithm in identifying significant biomarkers based on variable clustering. However, high dimensionality but small sample size of omics data makes self-organizing map-based model problematic in terms of selection stability and reproducibility. A novel feature screening system is presented in this study by coupling bootstrap with synergy self-organizing map-based orthogonal partial least squares discriminant analysis for stable and biologically meaningful metabolic biomarker selection. In the proposed feature screening system, particle swarm optimization algorithm is utilized to configure synergy self-organizing map-based orthogonal partial least squares discriminant analysis to perform the combination of clusters in a heuristic learning manner, enabling flexible selection of more informative features cost-effectively. Based on the paradigm of ensemble feature selection, bootstrap is adopted to explore significant variables consistently identified across multiple feature selectors rather than a single one. The feasibility of the novel feature screening system is evaluated by two most common inherited metabolic diseases, methylmalonic academia and propionic academia, using urinary metabolomics data. With the desirable classification performance, the proposed feature screening system outperforms simpler techniques in the identification of more features closely correlated with the metabolic mechanisms and the stability of selected candidate biomarkers against sample variations. Besides, the novel feature screening system greatly degrades the sensitivity of identified candidate biomarkers to the network size of self-organizing map, benefiting the identification of a suitable and stable final candidate biomarker list., (Copyright © 2020 Elsevier B.V. All rights reserved.)

This study compared the dosimetric and radiobiological parameters of prostate volumetric modulated arc therapy (VMAT) plans using different prescriptions optimized by the photon optimization (PO) and progressive resolution optimization (PRO) algorithm. A total of 20 prostate patients were selected retrospectively and divided into 2 groups of VMAT plans using prescriptions of 60 Gy/20 fx and 79 Gy/38 fx. Inverse treatment planning optimized by the PO and PRO algorithm based on the dual-arc technique was carried out by the Eclipse treatment planning system. The maximum dose, minimum dose, mean dose, dose-volume points, and dose-volume indices of the targets and organs at risk (OAR) were calculated from the plans. In addition, radiobiological parameters such as tumor control probability (TCP), normal tissue complication probability (NTCP), and equivalent uniform dose (EUD) of the targets and OAR were determined based on their dose-volume histograms (DVHs). A paired Student's t-test was carried out to compare the difference between mean dose-volume points, radiobiological parameters, and dose-volume indices. Two-tailed p < 0.05 was defined as having statistical difference. For prostate VMAT plans optimized by the PO algorithm, equal or slightly larger mean dose and TCP of the PTV (1% for 60 Gy/20 fx and 0.2% for 78 Gy/39 fx) were found by comparing to the PRO. These were followed by finding the slightly larger conformity index (CI; 0.927 vs 0.895 and 0.910 vs 0.904), larger or equal homogeneity index (HI; 0.054 vs 0.052 and 0.058 vs 0.058), and smaller gradient index (GI; 1.366 vs 2.288 and 1.585 vs 1.742) of the PTV using plans optimized by the PO vs PRO using prescriptions of 60 Gy/20 fx and 78 Gy/39 fx. For the OAR, we found that the mean doses, NTCPs, and EUDs of the rectum, bladder, and femur were slightly larger for plans optimized by the PO algorithm compared to the PRO, though both optimization algorithms satisfied all the dose-volume criteria and objectives in the inverse planning. Both the PO and PRO algorithm can generate prostate VMAT plans fulfilling the required dose-volume criteria. It is concluded that plans optimized by the PO algorithm can produce prostate plan with very similar quality compared to PRO., (Copyright © 2019 American Association of Medical Dosimetrists. Published by Elsevier Inc. All rights reserved.)

High-grade serous ovarian cancer (HGSOC) is one of the most common and lethal gynecological cancers. Long noncoding RNAs (lncRNAs) play important roles and act as prognostic biomarkers of ovarian cancer. However, few studies have focused on the prognostic prediction of lncRNAs solely in HGSOC. In this study, we identified candidate lncRNAs for a prognostic evaluation by examining reannotated lncRNA expression profiles and clinical data of 343 patients with HGSOC from The Cancer Genome Atlas. We built a 10-lncRNA signature using Cox-LASSO regression to predict the prognosis of patients with HGSOC. Trichotomized by the 10-lncRNA signature, high-risk patients experienced significantly shorter disease-free survival and overall survival (OS). Our novel 10-lncRNA signature showed superior predictive capacity compared to the other 2 published lncRNA signature models and clinicopathological parameters. We developed a nomogram for clinical use by integrating the 10-lncRNA signature and two clinicopathological risk factors to predict 1-, 3-, and 5-year OS. In addition, gene set enrichment analysis suggested that the group of high-risk patients was associated with mitotic spindle pathways. This model was also compatible with patients with or without BRCA1/2 mutations and had the potential to predict the response to platinum-based adjuvant chemotherapy. Our findings provide a novel 10-lncRNA prognostic signature for further clinical application in patients with HGSOC and indicate the underlying mechanisms of HGSOC progression.

This paper presents a generalized integrated framework of semi-automatic surgical template design. Several algorithms were implemented including the mesh segmentation, offset surface generation, collision detection, ruled surface generation, etc., and a special software named TemDesigner was developed. With a simple user interface, a customized template can be semi- automatically designed according to the preoperative plan. Firstly, mesh segmentation with signed scalar of vertex is utilized to partition the inner surface from the input surface mesh based on the indicated point loop. Then, the offset surface of the inner surface is obtained through contouring the distance field of the inner surface, and segmented to generate the outer surface. Ruled surface is employed to connect inner and outer surfaces. Finally, drilling tubes are generated according to the preoperative plan through collision detection and merging. It has been applied to the template design for various kinds of surgeries, including oral implantology, cervical pedicle screw insertion, iliosacral screw insertion and osteotomy, demonstrating the efficiency, functionality and generality of our method.

This paper presents a novel optimization algorithm called self-organizing potential field network (SOPFN). The SOPFN algorithm is derived from the idea of the vector potential field. In the proposed network, the neuron with the best weight is considered as the target with the attractive force, while the neuron with the worst weight is considered as the obstacle with the repulsive force. The competitive and cooperative behaviors of SOPFN provide a remarkable ability to escape from the local optimum. Simulations were performed, compared, and analyzed on eight benchmark functions. The results presented illustrate that the SOPFN algorithm achieves a significant performance improvement on multimodal problems compared with other evolutionary optimization algorithms.

Demodulation and decoding are pivotal for the eLoran system's timing and information transmission capabilities. This paper proposes a novel demodulation algorithm leveraging a multiclass support vector machine (MSVM) for pulse position modulation (PPM) of eLoran signals. Firstly, the existing demodulation method based on envelope phase detection (EPD) technology is reviewed, highlighting its limitations. Secondly, a detailed exposition of the MSVM algorithm is presented, demonstrating its theoretical foundations and comparative advantages over the traditional method and several other methods proposed in this study. Subsequently, through comprehensive experiments, the algorithm parameters are optimized, and the parallel comparison of different demodulation methods is carried out in various complex environments. The test results show that the MSVM algorithm is significantly superior to traditional methods and other kinds of machine learning algorithms in demodulation accuracy and stability, particularly in high-noise and -interference scenarios. This innovative algorithm not only broadens the design approach for eLoran receivers but also fully meets the high-precision timing service requirements of the eLoran system. [ABSTRACT FROM AUTHOR]

Objectives: Accurate contouring of the clinical target volume (CTV) is a key element of radiotherapy in cervical cancer. We validated a novel deep learning (DL)-based auto-segmentation algorithm for CTVs in cervical cancer called the three-channel adaptive auto-segmentation network (TCAS). Methods: A total of 107 cases were collected and contoured by senior radiation oncologists (ROs). Each case consisted of the following: (1) contrast-enhanced CT scan for positioning, (2) the related CTV, (3) multiple plain CT scans during treatment and (4) the related CTV. After registration between (1) and (3) for the same patient, the aligned image and CTV were generated. Method 1 is rigid registration, method 2 is deformable registration, and the aligned CTV is seen as the result. Method 3 is rigid registration and TCAS, method 4 is deformable registration and TCAS, and the result is generated by a DL-based method. Results: From the 107 cases, 15 pairs were selected as the test set. The dice similarity coefficient (DSC) of method 1 was 0.8155 ± 0.0368; the DSC of method 2 was 0.8277 ± 0.0315; the DSCs of method 3 and 4 were 0.8914 ± 0.0294 and 0.8921 ± 0.0231, respectively. The mean surface distance and Hausdorff distance of methods 3 and 4 were markedly better than those of method 1 and 2. Conclusions: The TCAS achieved comparable accuracy to the manual delineation performed by senior ROs and was significantly better than direct registration. [ABSTRACT FROM AUTHOR]

Pose estimation is the basis and key of human motion recognition. In the two-dimensional human pose estimation based on image, in order to reduce the adverse effects of mutual occlusion among multiple people and improve the accuracy of motion recognition, a structurally symmetrical two-dimensional multi-person pose estimation model combined with face detection is proposed in this paper. First, transfer learning is used to initialize each sub-branch network model. Then, MTCNN is used for face detection to predict the number of people in the image. According to the number of people, the image is input into the improved two-branch OpenPose network. What is more, the double judgment algorithm is proposed to correct the false detection of MTCNN. The experimental results show that compared with TensorPose, which is the latest improved method based on OpenPose, the Average Precision (AP) (Intersection over Union (IoU) = 0. 5) on the validation set is 8.8 higher. Furthermore, compared with OpenPose, the mean AP (IoU = 0. 5 : 0. 9 5) is 1.7 higher on the validation set and is 1.3 higher on the Test-dev test set. [ABSTRACT FROM AUTHOR]

Featured Application: The speckle binarization is the key process in binary ghost imaging and it can influence the imaging resolution and the contrast of the reconstruction results. Based on the finding in this study, the binary ghost imaging can reconstruct the image of target with high quality, which is beneficial to the application of binary ghost imaging in practice. The reconstruction quality of binary ghost imaging depends on the speckle binarization process. In order to obtain better binarization speckle and improve the reconstruction quality of binary ghost imaging, a local adaptive binarization method based on the fuzzy integral is proposed in this study. There are three steps in the proposed binarization process. The first step is to calculate the integral image of the speckle with the summed-area table algorithm. Secondly, the fuzzy integral image is calculated through the discrete Choquet integral. Finally, the binarization threshold of each pixel of the speckle is selected based on the calculated fuzzy integral result. The experiment results verify the feasibility of the proposed method. Compared with other methods qualitatively and quantitatively, the proposed method has high performance in reconstructing image of target. [ABSTRACT FROM AUTHOR]

Matching pursuit is able to decompose signals adaptively into a series of wavelets and has been widely applied in signal processing of the geophysical fields. Single-channel matching pursuit could not take into account the lateral continuity of seismic traces, and the recent multichannel matching pursuit exploits the lateral coherence as a constraint, which helps to improve the stability of decomposition results. However, atoms searched by multichannel matching pursuit currently are just shared by lateral seismic traces at the same time slicers. The lack of directionality in multichannel search strategies leads to irrationality in dealing with large dip angle seismic traces. Considering that the waveforms of reflection events are relatively continuous and similar, an improved multichannel matching pursuit is proposed to realize the directional decomposition of adjacent signals. Based on the principle of seismic reflection events tracking and identification, directional multichannel decomposition of seismic traces is realized. The seismic channel to be decomposed is correlated with the time shift of the optimal atom determined by the previous seismic channel. The time position of the maximum correlation indicates the center time of the optimal atom. Optimal atoms identified by one iteration of multichannel decomposition have the same frequency and phase parameters, different center time and amplitude parameters. The center time of the optimal atoms is consistent with seismic reflection events. Tests illustrate that the algorithm can successfully reconstruct 2D seismic data without reducing accuracy. Besides, the application of field data is of great significance for reservoir exploration and hydrocarbon interpretation. [ABSTRACT FROM AUTHOR]

In this paper, we propose a novel chaotic image encryption algorithm which involves a block image scrambling scheme and a new dynamic index based diffusion scheme. Firstly, the original image is divided into two equal blocks by vertical or horizontal directions. Then, we use the chaos matrix to construct X coordinate, Y coordinate and swapping control tables. By searching the X coordinate and Y coordinate tables, the swapping position of the processing pixel is located. The swapping control table is used to control the swapping of the pixel in the current block or the other block. Finally, the dynamic index scheme is applied to the diffusing of the scrambled image. The simulation results and performance analysis show that the proposed algorithm has an excellent safety performance with only one round. [ABSTRACT FROM AUTHOR]

Recently, a number of chaos-based image encryption algorithms have been proposed at the pixel level, but little research at the bit level has been conducted. This paper presents a novel bit-level image encryption algorithm that is based on piecewise linear chaotic maps (PWLCM). First, the plain image is transformed into two binary sequences of the same size. Second, a new diffusion strategy is introduced to diffuse the two sequences mutually. Then, we swap the binary elements in the two sequences by the control of a chaotic map, which can permute the bits in one bitplane into any other bitplane. The proposed algorithm has excellent encryption performance with only one round. The simulation results and performance analysis show that the proposed algorithm is both secure and reliable for image encryption. [ABSTRACT FROM AUTHOR]

Machine-to-machine (M2M) technology that allows machine devices to communicate directly with each other without human intervention has emerged as a cutting-edge technology for next-generation communication. As its applications become more and more popular, the number of users grows quickly and the quality of service needs to be improved. The traffic allocation problem may lead to many other problems such as low utilization ratio of the bandwidth and the jitter. When we tried to analyze the load on different paths using the simplest multipath routing algorithms, we found that the load changes just like simple harmonic motion, which inspires us to use physic and mathematic methods and tools to analyze the loading balance problems. And then, we designed a new traffic allocation algorithm. Compared with other algorithms, the algorithm proposed in this paper has much wider applicability. And the performance comparison experiment shows that DBTAs can really achieve improvement in decreasing the packet delay variance. [ABSTRACT FROM AUTHOR]

Abstract: A new Self-Organizing Map algorithm, called the probabilistic polar self-organizing map (PPoSOM), is proposed. PPoSOM is a new variant of PolSOM, which is constructed on 2-D polar coordinates. Two variables: radius and angle are used to reflect the data characteristics. PPoSOM, developed to enhance the visualization performance, provides more data characteristics compared with the traditional methods that use Euclidian distance as the only variable. The weight-updating rule of PPoSOM is associated with a cost function. Instead of using the hard assignment, PPoSOM employs the soft assignment that the assignment of data to neuron is based on a probabilistic function. The obtained results are compared with the conventional SOM and ViSOM. The presented results show that the proposed PPoSOM is an effective method for multidimensional data visualization. In addition, the quality measurement of mapping, synthetical cluster density (SCD) is applied and it shows PPoSOM exhibits an improved result compared with PolSOM. [Copyright &y& Elsevier]

Abstract: In this paper, a new algorithm named polar self-organizing map (PolSOM) is proposed. PolSOM is constructed on a 2-D polar map with two variables, radius and angle, which represent data weight and feature, respectively. Compared with the traditional algorithms projecting data on a Cartesian map by using the Euclidian distance as the only variable, PolSOM not only preserves the data topology and the inter-neuron distance, it also visualizes the differences among clusters in terms of weight and feature. In PolSOM, the visualization map is divided into tori and circular sectors by radial and angular coordinates, and neurons are set on the boundary intersections of circular sectors and tori as benchmarks to attract the data with the similar attributes. Every datum is projected on the map with the polar coordinates which are trained towards the winning neuron. As a result, similar data group together, and data characteristics are reflected by their positions on the map. The simulations and comparisons with Sammon''s mapping, SOM and ViSOM are provided based on four data sets. The results demonstrate the effectiveness of the PolSOM algorithm for multidimensional data visualization. [Copyright &y& Elsevier]

The range accuracy of three-dimensional (3D) ghost imaging is derived. Based on the derived range accuracy equation, the relationship between the slicing number and the range accuracy is analyzed and an optimum slicing number (OSN) is determined. According to the OSN, an improved 3D ghost imaging algorithm is proposed to increase the range accuracy. Experimental results indicate that the slicing number can affect the range accuracy significantly and the highest range accuracy can be achieved if the 3D ghost imaging system works with OSN. [ABSTRACT FROM AUTHOR]

The log–polar transform (LPT) is introduced into the star identification because of its rotation invariance. An improved autonomous star identification algorithm is proposed in this paper to avoid the circular shift of the feature vector and to reduce the time consumed in the star identification algorithm using LPT. In the proposed algorithm, the star pattern of the same navigation star remains unchanged when the stellar image is rotated, which makes it able to reduce the star identification time. The logarithmic values of the plane distances between the navigation and its neighbor stars are adopted to structure the feature vector of the navigation star, which enhances the robustness of star identification. In addition, some efforts are made to make it able to find the identification result with fewer comparisons, instead of searching the whole feature database. The simulation results demonstrate that the proposed algorithm can effectively accelerate the star identification. Moreover, the recognition rate and robustness by the proposed algorithm are better than those by the LPT algorithm and the modified grid algorithm. [ABSTRACT FROM AUTHOR]

Nursing factors have been found to be associated with a reduction in readmission rates. Nevertheless, few attentions have been given to the effect of nursing factors on nursing home (NH) residents. This study was to assess the impact of nursing factors on the hospital readmissions and length of stay (LOS) of the elderly residents from the NH. This was a cross-sectional study. Data were extracted from the NH of the LTCFocus.org data set between 2011 and 2018. The study included residents aged ≥55 years who were admitted to NH in the United States, following a hospitalization event. The nursing factors included facility-level data elements and medical care personnel. An unsupervised machine learning algorithm (K-means) was used to cluster NH according to readmission rate and LOS. Multivariate logistic regression analysis was performed. This study consisted of 107,000 NH-year observations. The median readmission rate was 17%, with a median LOS was 28.00 days. Three clusters were identified: cluster 1 was a high readmission rate with high LOS, cluster 2 was a low readmission rate with low LOS, and cluster 3 was a high readmission rate with low LOS. Multifacility and admission/bed were associated with a reduction in readmission rate and LOS in both cluster 1 vs cluster 2 and cluster 3 vs cluster 2. The special care unit and registered nurses' ratio were associated with decreased readmission rate and LOS in cluster 1 vs cluster 2. Total beds and Alzheimer unit decreased the readmission rate and LOS, whereas certified nursing assistant increased the readmission rate and LOS in cluster 3 vs cluster 2. NH for profit was associated with elevated readmission rate and LOS in cluster 1 vs cluster 2 and decreased readmission rate and LOS in cluster 3 vs cluster 2. Based on the subgroup analysis, the certified nursing assistant decreased readmission rate and LOS in cluster 1 vs cluster 2 and increased readmission rate and LOS in cluster 3 vs cluster 2 (all P < 0.005). This study indicates the importance of the improvement of nurse number and level and the inputs of facility characteristics in NH. [ABSTRACT FROM AUTHOR]

Community-acquired pneumonia (CAP) is one of the major global health problems worldwide. However, the epidemiological information of CAP is limited in China. This study aimed to estimate the incidence rate of CAP and describe the epidemiologic characteristics among the Chinese population. We conducted a retrospective analysis of CAP incidence using the Chinese Urban Basic Medical Insurance database of 23 provinces in 2016, which covered 427.52 million urban beneficiaries of all age groups in Mainland China. CAP episodes were identified using a diagnosis-term-derived algorithm, and multiple CAP records of one single person within 90 continuous days were considered as one single episode. The incidence rates were calculated and described by sex, age, region, and season. A total of 1.42 million patients were identified as having one or more CAP episodes, and finally a sum of 1.48 million CAP episodes were counted. The overall incidence of CAP was 7.13 (95% CI: 6.11–8.15) per 1000 person-years, in males 7.32 (95% CI: 6.28–8.35) and females 6.93 (95% CI: 5.92–7.94) per 1000 person-years, respectively. The incidence varied by age with a U-shaped curve peaking in children aged < 5 years old [65.80 (95% CI: 62.52–69.08)] and elderly population aged ≥ 80 years old [14.98 (95% CI: 13.63–16.34)]. The incidence varied markedly by regions. Furthermore, the rate showed a clear seasonal trend, which peaked in spring, decreased in summer and autumn, and re-ascended in winter. This study reveals a relatively high level of CAP incidence in China. These findings provide baseline data for establishing effective prevention strategies, targeted at susceptible populations, regions, and seasons in China. [ABSTRACT FROM AUTHOR]

Abstract: This paper demonstrates the application of near-infrared (NIR) process analysis to study gas–solid adsorption process non-invasively: its experimental setup, data treatment, and potentials as a convenient tool to investigate the gas–solid adsorption process. The experimental setup includes a differential adsorption bed (DAB) monitored by a NIR spectrometer via an optical fiber probe, which makes it convenient and reliable to construct adsorption mass-transfer models. A chemometrics strategy based on back propagation-artificial neural network (BP-ANN) and partial least squares (PLS) has been developed to treat NIR spectra collected during the adsorption process because of the obvious nonlinearity in concentration prediction. This nonlinear problem results from the great concentration variation of the adsorbate adsorbed by the adsorbent during the whole adsorption process, the extraordinarily low concentration of the adsorbed adsorbate at the beginning of the process, and probably NIR distinction between the adsorbate on the first adsorption layer at the beginning of the process and that on the other layers afterward. With the strategy, NIR spectra are pretreated with PLS for data compression and noise reduction, and then a BP-ANN is built as the nonlinear calibration model. As compared with linear calibration algorithm, our strategy has the higher predication ability for the whole adsorption process, even with less calibration samples. Finally, as an example the kinetics of aniline–silica gel adsorption process has been studied through the experimental setup and chemometrics strategy. [Copyright &y& Elsevier]

Abstract: Multivariate spectral analysis has been widely applied in chemistry and other fields. Spectral data consisting of measurements at hundreds and even thousands of analytical channels can now be obtained in a few seconds. It is widely accepted that before a multivariate regression model is built, a well-performed variable selection can be helpful to improve the predictive ability of the model. In this paper, the concept of traditional wavelength variable selection has been extended and the idea of variable weighting is incorporated into least-squares support vector machine (LS-SVM). A recently proposed global optimization method, particle swarm optimization (PSO) algorithm is used to search for the weights of variables and the hyper-parameters involved in LS-SVM optimizing the training of a calibration set and the prediction of an independent validation set. All the computation process of this method is automatic. Two real data sets are investigated and the results are compared those of PLS, uninformative variable elimination-PLS (UVE-PLS) and LS-SVM models to demonstrate the advantages of the proposed method. [Copyright &y& Elsevier]

Abstract: An improved method based on an ensemble of Monte Carlo uninformative variable elimination (EMCUVE) is presented for wavelength selection in multivariate calibration of spectral data. The proposed algorithm introduces Monte Carlo (MC) strategy to uninformative variable elimination-PLS (UVE-PLS) instead of leave-one-out strategy for estimating the contributions of each wavelength variable in the PLS model. In EMCUVE wavelength variables are evaluated by different Monte Carlo uninformative variable elimination (MCUVE) models. Moreover, a fusion of MCUVE and the vote rule can obtain an improvement over the original uninformative variable elimination method. Results obtained from simulated data and real data sets demonstrate that EMCUVE can properly carry out wavelength selection in the course of data analysis and improve predictive ability for multivariate calibration model. [Copyright &y& Elsevier]

Abstract: In the present study a new version of nonlinear partial least-square method based on artificial neural network transformation (ANN-NLPLS) has been proposed. This algorithm firstly transforms the training descriptors into the hidden layer outputs using the universal nonlinear mapping carried by an artificial neural network, and then utilizes PLS to relate the outputs of the hidden layer to the bioactivities. The weights between the input and hidden layers are optimized by a particle swarm optimization (PSO) method using the criterion of minimized model error via PLS modeling. An F-statistic is introduced to determine automatically the number of PLS components during the weight optimization. The performance is assessed using a simulated data set and two quantitative structure–activity relation (QSAR) data sets. Results of these three data sets demonstrate that ANN-NLPLS offers enhanced capacity in modeling nonlinearity while circumventing the overfitting frequently involved in nonlinear modeling. [Copyright &y& Elsevier]