Support Vector Machine Regression Model Research Articles

Ensemble Machine Learning Model to Predict the Waterborne Syndrome

The COVID-19 epidemic has highlighted the significance of sanitization and maintaining hygienic access to clean water to reduce mortality and morbidity cases worldwide. Diarrhea is one of the prevalent waterborne diseases caused due to contaminated water in many low-income countries with similar living conditions. According to the latest statistics from the World Health Organization (WHO), diarrhea is among the top five primary causes of death worldwide in low-income nations. The condition affects people in every age group due to a lack of proper water used for daily living. In this study, a stacking ensemble machine learning model was employed against traditional models to extract clinical knowledge for better understanding patients’ characteristics; disease prevalence; hygienic conditions; quality of water used for cooking, bathing, and toiletries; chemicals used; therapist’s medications; and symptoms that are reflected in the field study data. Results revealed that the ensemble model provides higher accuracy with 98.90% as part of training and testing phases when experimented against frequently used J48, Naïve Bayes, SVM, NN, PART, Random Forest, and Logistic Regression models. Managing outcomes of this research in the early stages could assist people in low-income countries to have a better lifestyle, fewer infections, and minimize expensive hospital visits.

Classification and Regression Models for Genomic Selection of Skewed Phenotypes: A Case for Disease Resistance in Winter Wheat (Triticum aestivum L.)

Most genomic prediction models are linear regression models that assume continuous and normally distributed phenotypes, but responses to diseases such as stripe rust (caused by Puccinia striiformis f. sp. tritici) are commonly recorded in ordinal scales and percentages. Disease severity (SEV) and infection type (IT) data in germplasm screening nurseries generally do not follow these assumptions. On this regard, researchers may ignore the lack of normality, transform the phenotypes, use generalized linear models, or use supervised learning algorithms and classification models with no restriction on the distribution of response variables, which are less sensitive when modeling ordinal scores. The goal of this research was to compare classification and regression genomic selection models for skewed phenotypes using stripe rust SEV and IT in winter wheat. We extensively compared both regression and classification prediction models using two training populations composed of breeding lines phenotyped in 4 years (2016–2018 and 2020) and a diversity panel phenotyped in 4 years (2013–2016). The prediction models used 19,861 genotyping-by-sequencing single-nucleotide polymorphism markers. Overall, square root transformed phenotypes using ridge regression best linear unbiased prediction and support vector machine regression models displayed the highest combination of accuracy and relative efficiency across the regression and classification models. Furthermore, a classification system based on support vector machine and ordinal Bayesian models with a 2-Class scale for SEV reached the highest class accuracy of 0.99. This study showed that breeders can use linear and non-parametric regression models within their own breeding lines over combined years to accurately predict skewed phenotypes.

Compressive Strength Prediction of Lightweight Concrete: Machine Learning Models

Concrete is the most commonly used construction material. The physical properties of concrete vary with the type of concrete, such as high and ultra-high-strength concrete, fibre-reinforced concrete, polymer-modified concrete, and lightweight concrete. The precise prediction of the properties of concrete is a problem due to the design code, which typically requires specific characteristics. The emergence of a new category of technology has motivated researchers to develop mechanical strength prediction models using Artificial Intelligence (AI). Empirical and statistical models have been extensively used. These models require a huge amount of laboratory data and still provide inaccurate results. Sometimes, these models cannot predict the properties of concrete due to complexity in the concrete mix design and curing conditions. To conquer such issues, AI models have been introduced as another approach for predicting the compressive strength and other properties of concrete. This article discusses machine learning algorithms, such as Gaussian Progress Regression (GPR), Support Vector Machine Regression (SVMR), Ensemble Learning (EL), and optimized GPR, SVMR, and EL, to predict the compressive strength of Lightweight Concrete (LWC). The simulation approaches of these trained models indicate that AI can provide accurate prediction models without undertaking extensive laboratory trials. Each model’s applicability and performance were rigorously reviewed and assessed. The findings revealed that the optimized GPR model (R = 0.9803) used in this study had the greatest accuracy. In addition, the optimized SVMR and GPR model showed good performance, with R-values 0.9777 and 0.9740, respectively. The proposed model is economic and efficient, and can be adopted by researchers and engineers to predict the compressive strength of LWC.

Determination of total flavonoid and polysaccharide content in Anoectochilus formosanus in response to different light qualities using hyperspectral imaging

This study demonstrated the capability of Vis/NIR hyperspectral imaging (400–1000 nm) in determining total flavonoid (TF) and total polysaccharide (TP) content in Anoectochilus formosanus that were cultivated in different light quality environments. We extracted spectra from the whole leaf (the control) and four zones (leaf tip, margin, middle and base) respectively. Partial least squares regression (PLSR) and support vector machine regression (SVMR) were used for building prediction models using the full spectral range. PLSR models outperformed SVMR models with determination of the coefficients of calibration (R2c) and validation (R2v) being over 0.9982 and the residual predictive deviations (RPD) being over 23.21. Comparing to the model for the control, the models for the four zones achieved similar performance; the model for the middle zone achieved slightly better performance than models for the other three zones (R2v = 0.9974, RMSEV = 0.14, RPD = 18.04 for TF; and R2v = 0.9982, RMSEV = 0.20, RPD = 23.56 for TP). We further selected effective wavelengths through the correlation coefficients between single wavelength, dual-wavelength combinations and measured content of TF and TP to simplify prediction models. New PLSR models were built based on the effective wavelengths with R2v, RMSEV and RPD of 0.8990, 0.85 and 3.07 for TF, and 0.8842, 1.71 and 2.79 for TP. Obtained results showed that hyperspectral imaging is promising for rapid and accurate determination of TF content and TP content in A. formosanus based on the PLSR models and the identified effective wavelengths.

Abstract P1-12-04: The impact of chemotherapy in elderly early triple negative breast cancer: A population based study from the SEER database

Abstract Introduction: The management of senile breast cancer is still controversial, especially chemotherapy. Therefore, we conduct an analysis of the impact of chemotherapy and establish prediction models of prognosis in elderly early TNBC by using machine learning. Methods: We enrolled 4,696 patients from SEER database according to the following criteria: female; year of diagnosis from 2010 to 2016; age of diagnosis ≥ 70 years; invasive ductal carcinoma as the only primary malignant cancer; stage I-III; TNBC subtype. The PSM method was utilized to reduce covariable imbalance. Univariable and multivariable analyses were used to compare BCSS and OS. KNN, CatBoost, decision tree, random forest, Gradient Boost, LightGBM, neural network models, SVM, XGBoost and logistic regression models were developed to predict the 5-year OS and BCSS for patients in the chemotherapy group. Results: Compared to matched patients with no-chemotherapy, multivariate analysis showed a better survival in matched patients with chemotherapy(BCSS:HR = 0.594,95%CI = 0.479-0.737,p < 0.001; OS:HR = 0.532,95%CI = 0.446-0.635,p < 0.001). Stratified analyses by age and stage was also conducted to discriminate the patients benefited from chemotherapy. Chemotherapy didn’t lowered the risk of cancer-specific mortality and all-cause mortality in stage I cohort (70-79 years, BCSS, HR=1.058, 95% CI=0.609-1.838, p=0.841; OS HR=0.721, 95% CI=0.478-1.086, p=0.117; 80+ years, BCSS, HR=0.210, 95% CI=0.005-8.920, p=0.414; OS HR=0.458, 95% CI=0.048-3.918, p=0.458). But patients diagnosed with stage II can benefit from chemotherapy (70-79 years, BCSS, HR=0.628, 95% CI=0.424-0.932, p=0.021; OS HR=0.538, 95% CI=0.386-0.749, p<0.001; 80+ years, BCSS, HR=0.413, 95% CI=0.230-0.744, p=0.003; OS HR=0.416, 95% CI=0.260-0.667, p<0.001). We observed similar phenomena in the stage III cohort. We further investigated the effects of tumor and nodal status in stage II cases between the two groups. We found that chemotherapy was a better prognostic indicator for patients with T2N0M0 and stage IIb (T2N0M0, BCSS, HR=0.420, 95% CI=0.261-0.675, p<0.001; OS HR=0.0.361, 95% CI=0.243-0.536, p<0.001; stage IIb, BCSS, HR=0.752, 95% CI=0.454-1.246, p=0.269; OS HR=0.640, 95% CI=0419-0.978, p=0.039), but not for T1N1M0 (BCSS, HR=0.778, 95% CI=0.249-2.432, p=0.666; OS HR=1.072, 95% CI=0.458-2.508, p=0.872). For patients with grade I and II, no statistical survival differences were identified between chemotherapy and no-chemotherapy patients(BCSS, HR=0.781, 95% CI=0.445-1.368, p=0.387; OS HR=0.802, 95% CI=0.512-1.256, p=0.335). While for patients with grade III, the chemotherapy patients demonstrated a better prognosis than no-chemotherapy patients (BCSS, HR=0.558, 95% CI=0.440-0.707, p<0.001; OS, HR=0.492, 95% CI=0.405-0.599, p<0.001). Ten models mentioned above were conducted to predict the 5-year OS and BCSS for patients in the chemotherapy group. On average, the accuracy was 0.885 on 5-year BCSS and 0.858 on 5-year OS. The average precision of the examined ten algorithms was 0.880 on 5-year BCSS and 0.862 on 5-year OS. Similarly, the average sensitivity was 0.981 on 5-year BCSS and 0.971 on 5-year OS. There was average F1 score of 0.932 on 5-year BCSS and 0.913 on 5-year OS. In terms of the AUC, the highest AUC was observed for the LightGBM. Considering all the parameters above, the LightGBM outperformed all other algorithms. Conclusion: For stage I, T1N1M0, grade I and grade II elderly early triple negative invasive ductal carcinoma patients, chemotherapy could not improve OS and BCSS. Therefore, de-escalation therapy might be appropriate for selected patients. The LightGBM is practical and trustful to predict the 5-year OS and BCSS for patients in the chemotherapy group. Citation Format: KaiYan Huang, Jie Zhang, YuShuai Yu, YuXiang Lin, ChuanGui Song. The impact of chemotherapy in elderly early triple negative breast cancer: A population based study from the SEER database [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P1-12-04.

Airborne imaging spectroscopy for assessing land-use effect on soil quality in drylands

Global population growth has resulted in land-use (LU) changes in many natural ecosystems, causing deterioration in the environmental conditions that affect soil quality. The effect of LU on soil quality is acute in water-limited systems that are characterized by insufficient availability of soil organic resources. Thus, the main objective of this study was to assess the effects of human activities (i.e., land-uses as grazing, modern agriculture, and runoff harvesting systems) on soil quality using imaging spectroscopy (IS) in the arid regions of Israel. For this, 12 physical, biological, and chemical soil properties were selected and further integrated into the soil quality index (SQI) as a method to assess the significant effects of LU changes in an arid area in southern Israel. A flight campaign of the AisaFENIX hyperspectral airborne sensor was used to develop an IS prediction model for the SQI on a regional scale. The spectral signatures, extracted from the hyperspectral image itself, were well separable among the four LUs using the partial least squares-discriminant analysis (PLS-DA) classification method (OA = 95.31%, Kc = 0.90). The correlation was performed using multivariate support vector machine-regression (SVM-R) models between the spectral data and the measured soil indicators and the overall SQI. The SVM-R models were significantly correlated for several soil properties, including the overall SQI (R2adjVal = 0.87), with the successful prediction of the regional SQI mapping (R2adjPred = 0.78). Seven individual soil properties, including fractional sand and clay, SOM, pH, EC, SAR, and P, were successfully used for developing prediction maps. Applying IS, and statistically integrative methods for comprehensive soil quality assessments enhances the prediction accuracy for monitoring soil health and evaluating degradation processes in arid environments. This study establishes a precise tool for sustainable and efficient land management and could be an example for future potential IS earth-observing space missions for soil quality assessment studies and applications.

Carbon sink price prediction based on radial basis kernel function support vector machine regression model

Abstract At present, the main pricing methods for carbon sink trading are marginal cost pricing, shadow price pricing and physical option pricing. The most serious defect of these three methods is that their theoretical basis is the extremely idealized effective market hypothesis, which obviously does not conform to the fractal and chaotic behavior characteristics of the actual carbon sink trading market. It is particularly important to study new pricing technologies and methods for a complex, nonlinear and dissipative system. Therefore, this paper constructs a carbon sink price prediction method based on the radial basis kernel function support vector machine (RBF-SVM) model. The main findings are the following: • Compared with the build nonlinear support vector machine (SVM) models by the three kernel functions: linear kernel function (LF), radial basis function (RBF) and sigmoid kernel function, the RBF-SVM has the highest prediction accuracy. • The adaptability of the RBF-SVM model is tested by using the price fluctuation data of EU allowance (EUA). It is proved that the adaptability of the RBF-SVM model for the price predictions of heterogeneous carbon products is also excellent. • The forecast results of the RBF-SVM model for carbon sink price for the period of September 2021 to August 2022 show that the lowest carbon sink price of 20 euros is expected to occur around November 2021 and the highest price of 38.5 euros around August 2022. Therefore, the established RBF-SVM model can be used for carbon sink trading market price prediction, which is more scientific and applicable than the previous three methods.

Machine learning regression models for prediction of multiple ionospheric parameters

• Machine learning regression models to predict multiple ionospheric parameters. • Forecasting of f0F2, hmF2 and TEC parameters up to 1 h in the mid-latitude region. • Performance comparison of DT, RF and SVM regression models. • A novel approach as a dataset of digisonde and GNSS stations in close proximity. The variation of the ionospheric parameters has a crucial role in space weather, communication, and navigation applications. In this research, we analyze the prediction performance of three machine learning (ML) regression models, decision trees, random forest and support vector machine algorithms for the F2-layer critical frequency (f0F2), F2-layer height of the peak electron density (hmF2), and total electron content (TEC). The hourly f0F2 and hmF2 values of ROME (RO041) digisonde and hourly TEC values of a close by International GNSS Service (IGS) station (site code M0SE00ITA) were obtained for the period between January 1, 2012 and 31 December 2013. The inputs to be trained in the proposed methods are the observation periods of the data (sine and cosine of the day of year and hour of the day), hourly values of solar index F10.7 and geomagnetic index Ap, the present values of f0F2(t), hmF2(t), TEC(t), and their values at t − 23 h. The outputs are the predicted values of f0F2, hmF2, TEC at t + 1. The 2012 values of these parameters were used to train the models and they were predicted 1 h in advance during 2013. The root mean square error (RMSE) values between observed and predicted data were compared for the whole year, summer, winter, and equinox periods. The results showed that proposed ML regression models were successful in the prediction of multiple ionospheric parameters, but random forest regression provides more accurate prediction results than support vector machine and decision trees.

Bacterial prediction using internet of things (IoT) and machine learning.

Water is a basic and primary resource which is required for sustenance of life on the Earth. The importance of water quality is increasing with the ascending water pollution owing to industrialization and depletion of fresh water sources. The countries having low control on reducing water pollution are likely to retain poor public health. Additionally, the methods being used in most developing countries are not effective and are based more on human intervention than on technological and automated solutions. Typically, most of the water samples and related data are monitored and tested in laboratories, which eventually consumes time and effort at the expense of producing fewer reliable results. In view of the above, there is an imperative need to devise a proper and systematic system to regularly monitor and manage the quality of water resources to arrest the related issues. Towards such ends, Internet of Things (IoT) is a great alternative to such traditional approaches which are complex and ineffective and it allows taking remote measurements in real-time with minimal human involvement. The proposed system consists of various water quality measuring nodes encompassing various sensors including dissolved oxygen, turbidity, pH level, water temperature, and total dissolved solids. These sensors nodes deployed at various sites of the study area transmit data to the server for processing and analysis using GSM modules. The data collected over months is used for water quality classification using water quality indices and for bacterial prediction by employing machine learning algorithms. For data visualization, a Web portal is developed which consists of a dashboard of Web services to display the heat maps and other related info-graphics. The real-time water quality data is collected using IoT nodes and the historic data is acquired from the Rawal Lake Filtration Plant. Several machine learning algorithms including neural networks (NN), convolutional neural networks (CNN), ridge regression (RR), support vector machines (SVM), decision tree regression (DTR), Bayesian regression (BR), and an ensemble of all models are trained for fecal coliform bacterial prediction, where SVM and Bayesian regression models have shown the optimal performance with mean squared error (MSE) of 0.35575 and 0.39566 respectively. The proposed system provides an alternative and more convenient solution for bacterial prediction, which otherwise is done manually in labs and is an expensive and time-consuming approach. In addition to this, it offers several other advantages including remote monitoring, ease of scalability, real-time status of water quality, and a portable hardware.

Inversion of Soil Salinity Using Multisource Remote Sensing Data and Particle Swarm Machine Learning Models in Keriya Oasis, Northwestern China

Soil salinization is a global problem that damages soil ecology and affects agricultural development. Timely management and monitoring of soil salinity are essential to achieve the most sustainable development goals in arid and semi-arid regions. It has been demonstrated that Polarimetric Synthetic Aperture Radar (PolSAR) data have a high sensitivity to the soil dielectric constant and soil surface roughness, thus having great potential for the detection of soil salinity. However, studies combining PALSAR-2 data and Landsat 8 data to invert soil salinity information are less common. The particle swarm optimization (PSO) algorithm is characterized by simple operation, fast computation, and good adaptability, but there are relatively few studies applying it to soil salinity as well. This paper takes the Keriya Oasis as an example, proposing the PSO-SVR and PSO-BPNN models by combining PSO with support vector machine regression (SVR) and back-propagation neural network (BPNN) models. Then, PALSAR-2 data, Landsat 8 data, evapotranspiration data, groundwater burial depth data, and DEM data were combined to conduct the inversion study of soil salinity in the study area. The results showed that the introduction of PSO generated a satisfactory estimating performance. The SVR model accuracy (R2) improved by 0.07 (PALSAR-2 data), 0.20 (Landsat 8 data), and 0.19 (PALSAR + Landsat data); the BP model accuracy (R2) improved by 0.03 (PALSAR-2 data), 0.24 (Landsat 8 data), and 0.12 (PALSAR + Landsat data), and then combined with the model inversion plots, we found that PALSAR + Landsat data combined with the PSO-SVR model could achieve better inversion results. The fine texture information of PALSAR-2 data can be used to better invert the soil salinity in the study area by combining it with the rich spectral information of Landsat 8 data. This study complements the research ideas and methods for soil salinization using multi-source remote sensing data to provide scientific support for salinity monitoring in the study area.

Longitudinal resting-state functional connectivity and regional brain atrophy-based biomarkers of preclinical cognitive impairment in healthy old adults.

Intervention against age-related neurodegenerative diseases may be difficult once extensive structural and functional deteriorations have already occurred in the brain. Investigating 6-year longitudinal changes and implications of regional brain atrophy and functional connectivity in the triple-network model as biomarkers of preclinical cognitive impairment in healthy aging. We acquired longitudinal cognitive scores and magnetic resonance imaging (MRI) data from 74 healthy old adults. Resting-state functional MRI (rs-fMRI) analysis was conducted using FSL6.0.1 to examine functional connectivity changes and regional brain morphometries were quantified using FreeSurfer5.3. Finally, we cross-validated and compared two support vector machine (SVM) regression models to predict future 6-year cognition score from the baseline regional brain atrophy and resting-state functional connectivity (rs-FC) measures. After a 6-year follow-up, our results (P < 0.05-corrected) indicated significant connectivity reduction within all the three brain networks, significant differences in regional brain volumes and cortical thickness. We also observed significant improvement in episodic memory and significant decline in executive functions. Finally, comparing the two models, we observed that regional brain atrophy predictors were more efficient in approximating future 6-year cognitive scores (R = 0.756, P < 0.0001) than rs-FC predictors (R = 0.6, P < 0.0001). This study used longitudinal data to keep subject variability low and to increase the validity of the results. We demonstrated significant changes in structural and functional MRI over 6years. Our findings present a potential neuroimaging-based biomarker to detect cognitive impairment and prevent risks of neurodegenerative diseases in healthy old adults.

Data-Driven Modeling and Prediction Analysis for Surface Roughness of Special-Shaped Stone by Robotic Grinding

This paper aims to accurately predict the surface quality of the special-shaped stone by robotic grinding and effectively guide the adjustment of process parameters to ensure stable grinding quality, applies a support vector machine model based on improved whale optimization algorithm (IWOA-SVR), so as to establish a prediction model of surface roughness of special-shaped stone and a selection method of process parameters. The proposed IWOA-SVR was used to improve the prediction accuracy of support vector machine regression model, and a prediction model of surface roughness (Ra) for stone was established. On this basis, the relationship between the output parameters (surface roughness) and inutput parameters (spindle speed, feed speed, cutting depth and cutting width)was explored to obtain more suitable process parameters. Combining the grinding experiment data of special-shaped stone, the comparison was carried out between IWOA-SVR and the SVR model optimized by the commonly used optimization algorithms (grid-search optimization algorithm (GS) and whale optimization algorithm (WOA)). Under the same sample condition, the prediction error of GS-SVR is the most large, and the average prediction error of IWOA-SVR is only 86.1% if that of WOA-SVR, the training time is shortened by 54.4%. The influence of process parameters on surface roughness obtained by IWOA-SVR can effectively guide the selection and adjustment of process parameters. It has good guiding significance for maintaining the excellent grinding quality of special-shaped stone.

Gaussian Regression Models for Evaluation of Network Lifetime and Cluster-Head Selection in Wireless Sensor Devices

The paper presents a model predictive approach for evaluating network lifetime and cluster head selection for a wireless sensor network. The dynamic parameters of a wireless sensor network are collected using Smart Mesh IP Power and performance calculator. The study considers a machine learning approach to combine clustering with the optimal routing protocol. The hop depth, advertising, number of Motes, backbone, routing, reporting interval, payload size, downstream frame size, supply voltage, and path stability are the predictors, and the current consumption, data latency, and build time are the response variables to establish the models for estimating the power and performance of the network. The remaining energy in each node, distance from the base station, and data transmission rate are the predictors, and the priority of the cluster head is the response variable to establish models for achieving an optimal routing path in a wireless sensor network. The standard tree, Support Vector Machine, Ensemble, and Gaussian process regression models for lifetime estimation are analyzed in comparison with the Smart Mesh IP tool, and the models for cluster head selection are investigated in comparison with ANFIS based models. This novel approach concentrates on the effect of various dynamic parameters on network lifetime prediction.

Deep Learning Based Automated Detection of Diseases from Apple Leaf Images

In Agriculture Sciences, detection of diseases is one of the most challenging tasks. The mis-interpretations of plant diseases often lead to wrong pesticide selection, resulting in damage of crops. Hence, the automatic recognition of the diseases at earlier stages is important as well as economical for better quality and quantity of fruits. Computer aided detection (CAD) has proven as a supportive tool for disease detection and classification, thus allowing the identification of diseases and reducing the rate of degradation of fruit quality. In this research work, a model based on convolutional neural network with 19 convolutional layers has been proposed for effective and accurate classification of Marsonina Coronaria and Apple Scab diseases from apple leaves. For this, a database of 50,000 images has been acquired by collecting images of leaves from apple farms of Himachal Pradesh (H.P) and Uttarakhand (India). An augmentation technique has been performed on the dataset to increase the number of images for increasing the accuracy. The performance analysis of the proposed model has been compared with the new two Convolutional Neural Network (CNN) models having 8 and 9 layers respectively. The proposed model has also been compared with the standard machine learning classifiers like support vector machine, k-Nearest Neighbour, Random Forest and Logistic Regression models. From experimental results, it has been observed that the proposed model has outperformed the other CNN based models and machine learning models with an accuracy of 99.2%.

Empirical Analysis for Remaining Useful Life Estimation via Data-Driven Models

Today, due to the growing complexity of engineered systems, it is crucial to develop technologies able to deal with the systems' behavior to maintain a high degree of safety, reliability, and efficiency while reducing operating expenses such as maintenance costs. The idea is to develop a Prognostics and Health Management (PHM) framework to monitor these complex systems' behavior using sensory data and then apply machine learning models to infer the current health state. One important goal is to estimate the Remaining Useful Life (RUL) essential for optimizing maintenance processes and sustainable practices in industrial settings. This work presents an empirical analysis for RUL estimation via a model degradation built using condition monitoring data. A support vector machine regression (SVR) model and a similarity measure algorithm are employed to extract degradation trends and compute the RUL. We evaluate the prognostics performance and compare the results with reported benchmarks from publicized works.

The Influence of Different Ultrasonic Machines on Radiomics Models in Prediction Lymph Node Metastasis for Patients with Cervical Cancer.

Objective To investigate the effects of different ultrasonic machines on the performance of radiomics models using ultrasound (US) images in the prediction of lymph node metastasis (LNM) for patients with cervical cancer (CC) preoperatively. Methods A total of 536 CC patients with confirmed histological characteristics and lymph node status after radical hysterectomy and pelvic lymphadenectomy were enrolled. Radiomics features were extracted and selected with US images acquired with ATL HDI5000, Voluson E8, MyLab classC, ACUSON S2000, and HI VISION Preirus to build radiomics models for LNM prediction using support vector machine (SVM) and logistic regression, respectively. Results There were 148 patients (training vs validation: 102:46) scanned in machine HDI5000, 75 patients (53:22) in machine Voluson E8, 100 patients (69:31) in machine MyLab classC, 110 patients (76:34) in machine ACUSON S2000, and 103 patients (73:30) in machine HI VISION Preirus, respectively. Few radiomics features were reproducible among different machines. The area under the curves (AUCs) ranged from 0.75 to 0.86, 0.73 to 0.86 in the training cohorts, and from 0.71 to 0.82, 0.70 to 0.80 in the validation cohorts for SVM and logistic regression models, respectively. The highest difference in AUCs for different machines reaches 17.8% and 15.5% in the training and validation cohorts, respectively. Conclusions The performance of radiomics model is dependent on the type of scanner. The problem of scanner dependency on radiomics features should be considered, and their effects should be minimized in future studies for US images.

Estimating the frequency and bandwidth of square-split ring resonator (S-SRR) designs via support vector machines (SVM)

In this study Support Vector Machine (SVM) based estimation technique is proposed for estimating the bandstop frequency and bandwidth of square-split ring resonators. Artificially engineered surfaces especially the planar frequency selective surfaces like the SRRs have narrowband properties so that estimating the filtering frequencies and the bandwidth is essential in a cost and design-effective way. The proposed method, which is superior to optimization methods and 3D electromagnetic solvers in terms of cost and computational burden, achieved accurate results via SVMs generalization capability. This study represents two SVM regression models one for predicting frequency and the other for predicting bandwidth having fast response and accuracy. Results of the proposed model reveal that resonance frequency estimation error, in terms of percentage, is bounded in the interval [0.0542, 3.5938], with an overall error of 0.89 % for the test data. The mean and standard deviation of the percentage error is obtained as 0.9861 and 0.9376, respectively. In addition to that -10dB bandwidth is estimated with the bounded error where estimation error in terms of percentage would be lie in the interval [0.068925, 6.876800] with an overall error of 3.68% for the test data.

Projection of High-Dimensional Genome-Wide Expression on SOM Transcriptome Landscapes

The self-organizing maps portraying has been proven to be a powerful approach for analysis of transcriptomic, genomic, epigenetic, single-cell, and pathway-level data as well as for “multi-omic” integrative analyses. However, the SOM method has a major disadvantage: it requires the retraining of the entire dataset once a new sample is added, which can be resource- and time-demanding. It also shifts the gene landscape, thus complicating the interpretation and comparison of results. To overcome this issue, we have developed two approaches of transfer learning that allow for extending SOM space with new samples, meanwhile preserving its intrinsic structure. The extension SOM (exSOM) approach is based on adding secondary data to the existing SOM space by “meta-gene adaptation”, while supervised SOM portrayal (supSOM) adds support vector machine regression model on top of the original SOM algorithm to “predict” the portrait of a new sample. Both methods have been shown to accurately combine existing and new data. With simulated data, exSOM outperforms supSOM for accuracy, while supSOM significantly reduces the computing time and outperforms exSOM for this parameter. Analysis of real datasets demonstrated the validity of the projection methods with independent datasets mapped on existing SOM space. Moreover, both methods well handle the projection of samples with new characteristics that were not present in training datasets.

The utility of a near-infrared spectrometer to predict the maturity of green peas (Pisum sativum)

This work evaluated the utility of an embedded diode array near-infrared reflectance spectrometer to predict the hardness (maturity) of green peas. Partial least squares, support vector machine (SVM) and artificial neural network regression models were developed to predict tenderometer reading (TR) from near-infrared spectra over the range of maturities experienced throughout a harvest season. The SVM model achieved the best performance in an independent data set collected in a second harvest season. Here, the SVM model explained 83% of the variation in TR with a root mean standard error of prediction (RMSEP) of 4.4 TR units over a range of TR from 65 to 104. The RPD of 2.3 achieved in this study exceeds the threshold for rough screening but falls short of replacing the tenderometer for quality control.

Discovery of DNA-Carbon Nanotube Sensors for Serotonin with Machine Learning and Near-infrared Fluorescence Spectroscopy.

DNA-wrapped single walled carbon nanotube (SWNT) conjugates have distinct optical properties leading to their use in biosensing and imaging applications. A critical limitation in the development of DNA-SWNT sensors is the current inability to predict unique DNA sequences that confer a strong analyte-specific optical response to these sensors. Here, near-infrared (nIR) fluorescence response data sets for ∼100 DNA-SWNT conjugates, narrowed down by a selective evolution protocol starting from a pool of ∼1010 unique DNA-SWNT candidates, are used to train machine learning (ML) models to predict DNA sequences with strong optical response to neurotransmitter serotonin. First, classifier models based on convolutional neural networks (CNN) are trained on sequence features to classify DNA ligands as either high response or low response to serotonin. Second, support vector machine (SVM) regression models are trained to predict relative optical response values for DNA sequences. Finally, we demonstrate with validation experiments that integrating the predictions of ensembles of the highest quality neural network classifiers (convolutional or artificial) and SVM regression models leads to the best predictions of both high and low response sequences. With our ML approaches, we discovered five DNA-SWNT sensors with higher fluorescence intensity response to serotonin than obtained previously. Overall, the explored ML approaches, shown to predict useful DNA sequences, can be used for discovery of DNA-based sensors and nanobiotechnologies.