Multivariate Algorithm Research Articles

The Use of Some Nonlinear Functions to Explain Growth in Japanese Quails with Multivariate Adaptive Regression Splines Algorithm

The study aimed was to determine the best nonlinear function describing the growth stages of the Japanese quail breed. To this aim, growth functions such as exponential, logistic, von Bertalanffy, Brody, and Gompertz were used as nonlinear functions is used in the description of the body weight-age relationship of male and female Japanese quails. The Multivariate Adaptive Regression Splines (MARS) data mining algorithm was applied to the individual growth parameters obtained from the determined as the best fit model, and the relationship between sex and growth parameters with it has been revealed. The study dataset was 1267 body weight-age records collected from the hatching to the 6th week of age of 181 Japanese quails consisting of 90 females and 91 males. Each model was applied separately for both males and females. Model fit criteria such as coefficient of determination (R2), adjusted coefficient of determination (R2adj), Akaike's information criterion (AIC), and Bayes information criterion (BIC) were used to evaluate the performances of the growth functions used individually. All the statistical analyses were made by the R package program. The growth curve models were ranked in the form of Logistic > Gompertz > von Bertalanffy > Brody > Exponential according to the goodness of fit criteria. The most suitable model among the non-linear models in terms of performance was logistic. When the relationship between the growth curve parameters and body weight of the logistic model was explained with the MARS algorithm, the goodness of fit criteria showed that the obtained MARS model showed reliable performance. In addition, Pearson’s correlation coefficient between real and estimated body weight was found quite strong for the MARS algorithm (r=0.935). The results showed that the MARS algorithm can be presented as a good reference for breeders to establish breed standards and selection strategies for Japanese quails in growth parameters for breeding purposes.

Liquidity Pricing Risk and Crude Oil Market: Analyzing the Liquidity as a Priced Factor in Yields during the Pandemic Uncertainty

The liquidity risk can impact the transaction execution with the significant losses. In the crude oil market, this work examines the market liquidity as a priced risk factor in yields. This area may considerably matter to be investigated, while an unprecedented slump is observed in the crude oil price during the pandemic period. The analysis is derived through the time series and multivariate algorithm methods. Pre-pandemic uncertainty and in the era of easing coronavirus-related restrictions, the liquidity was not priced in returns of the same trading session. In the period of economic restrictions, the crude oil trading was exposed to the liquidity risk for the same trading session. This risk was compensated by a premium return in environments of price uncertainty. In this debate, the liquidity risk was a priced risk factor in yields. Evidence from the VAR findings reports, that the past time series of the liquidity was priced in returns before the pandemic crisis takes place. Post-easing economic restrictions and in the era of pandemic-related restrictions, the past times series of the liquidity was not priced in returns. This quantification of the market liquidity may be more applicable for the liquidity risk management in the crude oil market.

Thermodynamic design and optimization of pumped thermal electricity storage systems using supercritical carbon dioxide as the working fluid

Pumped thermal electricity storage systems are a potential approach to large-scale energy storage, and supercritical carbon dioxide (SCO2) is a promising working fluid. Therefore, this study designed a SCO2 pumped thermal electricity storage system based on the reversible Brayton cycle and clarified the characteristics and restrictions of using SCO2 as the working fluid. Then, a series of improved variants that included recuperation, the elimination of medium–low temperature heat storage, and recompression were developed in a targeted manner. Roundtrip efficiency and energy density were defined as the performance indicators of different configurations, and multi-parameters were optimized by using the Find minimum of constrained nonlinear multivariable function algorithm. This study focused on the influence of the heat transfer matching of SCO2 and storage medium on system performance. Results showed that the improved configuration using recompression had the highest roundtrip efficiency of 69.38%, which was 33.48%-pts and 14.11%-pts higher than the roundtrip efficiencies of the basic configuration without recuperation and the configuration with simple recuperation, respectively. Exergy analysis was carried out to quantify the exergy losses of components further. The maximum portion of exergy was destroyed in the heat exchangers between SCO2 and the storage medium in the nonrecuperative system, and in the recuperators in the recuperative systems. Therefore, improving the heat transfer matching of SCO2 with the storage medium and recuperator performance is the key to increasing efficiency. Finally, different high temperature heat storage materials for the SCO2 pumped thermal electricity system were compared, and suggestions for the selection of storage medium were provided.

Classification of Aggregates Using Multispectral Two-Dimensional Angular Light Scattering Simulations.

Airborne particulate matter plays an important role in climate change and health impacts, and is generally irregularly shaped and/or forms agglomerates. These particles may be characterized through their light scattering signals. Two-dimensional angular scattering from such particles produce a speckle pattern that is influenced by their morphology (shape and material composition). In what follows, we revisit morphological descriptors obtained from computationally generated light scattering patterns from aggregates of spherical particles. These descriptors are used as inputs to a multivariate statistical algorithm and then classified via supervised machine learning algorithms. The classification results show improved accuracy over previous efforts and demonstrate the utility of the proposed morphological descriptors.

End-to-end dataflow engineering framework of honey manufacturing from intermediates to process by TAS1R2@AuNPs/SPCE biosensor coupled with quality transfer principle

This study reported an original end-to-end dataflow engineering framework for the quality transfer principle to overcome the quality challenges in real-world honey manufacturing. Firstly, 650 pivotal data points of physical and chemical quality attributes from 65 batches of honey intermediates were characterized through multiple sensors, which included rheological properties, acidity, moisture, and sugars. Furthermore, a hypersensitized TAS1R2@AuNPs/SPCE biosensor was developed to identify biological quality attributes of honey, the powerful affinities between honey intermediates and the TAS1R2 receptor were discovered (KD < 1 × 10−8 M), and the abnormal batches of B2, B23 and C23 were diagnosed by TAS1R2@AuNPs/SPCE biosensor and multivariable algorithm. Finally, the end-to-end dataflow containing physical, chemical and biological critical quality attributes was successfully established to interpret the quality transfer principle of honey manufacturing, which revealed that the front-end refining process was relatively unstable and the back-end refining process was a negligible influence on the quality of honey manufacturing. This framework embraces quality management, quality transfer, and biosensor information, which will contribute to discovering the quality transfer principle in industrial innovation for intelligent manufacturing.

PM2.5 concentration forecasting through a novel multi-scale ensemble learning approach considering intercity synergy

Accurate PM 2.5 concentration prediction can provide reliable air pollution warning information to the public. However, previous studies have often focused on the data of the target city itself, ignoring the interaction among cities in the same region. In this paper, we develop a multi-scale ensemble learning approach to forecast daily PM 2.5 concentrations of the target city by modeling its air and climate indicators, and PM 2.5 value of its neighboring cities. First, the proposed approach smooths the multivariate data by singular spectrum analysis and performs multi-feature selection based on distance factor and predictive power of data. Second, the inherent association among the obtained multiple features is captured by multivariate empirical modal decomposition. Third, the Hurst exponent is applied to match each time scale with the corresponding predictor for multi-step prediction. Finally, the forecasting values of all time scales are summed to obtain the PM 2.5 concentration forecasting results of the target city. Four experiments involving Beijing, Wuhan, and Shenzhen are carried out to verify the accuracy and robustness of the proposed approach. The experimental results show that our approach outperforms all benchmark models, and introducing city synergy strategy can improve the forecasting performance significantly. • The synergistic effect of surrounding cities improves the forecasting accuracy. • Double-level feature selection is established to extract the effective features. • Multivariate decomposition algorithm implements multi-source data fusion well. • A forecasting model matching strategy is developed based on the average Hurst exponent. • Average MAPE improvements of the approach against benchmark models are above 8.9%.

Simulation and Optimization of Ultrasonic Transducer

The ultrasonic transducers have numerous applications in industries, including medical probes for performing ultrasound scans. One of the significant drawbacks of the ultrasonic transducer is the wastage of a large portion of energy, due to high acoustic impedance, while transmitting ultrasonic waves to the target object. The present study is aimed to investigate the material design of the piezo-composite transducer and improve its performance. Different piezo-composite transducers were simulated in the COMSOL environment by varying input parameters, and three key performance indicators (KPI) were calculated. Many constraint-based multivariable optimization algorithms have been used to maximize the KPIs. A set of parameters, such as Sensitivity and Fractional Bandwidth, have been found to increase the performance of piezo-composite transducer model and its overall efficiency. This study is intended to impinge unidirectional property to the transducer which is found to be beneficial in more accurate medical as well as structural reports and cost savings.

Design and Development of a Solar Electric Delivery Pod

The ultrasonic transducers have numerous applications in industries, including medical probes for performing ultrasound scans. One of the significant drawbacks of the ultrasonic transducer is the wastage of a large portion of energy, due to high acoustic impedance, while transmitting ultrasonic waves to the target object. The present study is aimed to investigate the material design of the piezo-composite transducer and improve its performance. Different piezo-composite transducers were simulated in the COMSOL environment by varying input parameters, and three key performance indicators (KPI) were calculated. Many constraint-based multivariable optimization algorithms have been used to maximize the KPIs. A set of parameters, such as Sensitivity and Fractional Bandwidth, have been found to increase the performance of piezo-composite transducer model and its overall efficiency. This study is intended to impinge unidirectional property to the transducer which is found to be beneficial in more accurate medical as well as structural reports and cost savings.

20 Outcome Study of Mild Traumatic Brain Injury Patients Integrating a Brain Electrical Activity-Based Decision Rule

Clinical decision rules such as the Canadian Head CT rule have high sensitivity but lack specificity for identifying significant intracranial findings when evaluating patients with mild traumatic brain injury (mTBI). Advances in brain electrical activity (EEG) signal processing, real-time analyses, and use of an AI/machine learning for the derivation of brain activity-based biomarkers have greatly enhanced the pragmatism of EEG clinically. High accuracy and negative predictive value have been demonstrated (n-720) using an FDA cleared brain activity-based multivariate algorithm for predicting the likelihood of intracranial injuries with ≥ 1mL blood visible on a CT scan.

Array‐Based Discrimination of Phenolic Acids in Water at Near‐Neutral pH with Chromogenic and Fluorescent Cyclodextrin‐Based Supramolecular Sensors

AbstractPhenolic acids are a class of poorly water‐soluble organic compounds with antioxidant properties that contain a carboxylic acid and a phenol group, common in plants. We used hydroxypropyl‐β‐cyclodextrin (HP‐β‐CD) as a supramolecular host to sense phenolic acids in water near physiological pH. The analytes’ complexation was monitored through changes in the optical absorption and emission properties of a series of six organic dyes, acting as indicators in a displacement assay, that were selected from screening a panel of 17 candidates. Binding constants between these dyes and the HP‐β‐CD were in the 102–104 range. We showed that the nuanced differences in the dyes’ optical signatures associated with this displacement process, when analyzed and summarized using multivariate analysis algorithms such as principal component analysis (PCA), can be used to successfully discriminate among six phenolic acids, and to identify six unknown samples of such acids, down to a 0.02 mM concentration.

Robust multimodel control for uncertain nonlinear MIMO systems based on ARX-Laguerre multimodel and LSDP approach

In this paper, we propose a new alternative for the robust control of nonlinear MIMO (multiple-input multiple-output) systems based on the multimodel approach and LSDP (Loop-Shaping Design Procedure)approach in the discrete case. First, the multimodel approach is exploited by representing each linear sub-model in the form of the recent linear MIMO ARX (Auto-Regressive with eXogenous inputs) -Laguerre model which guarantees a significant parameters number reduction compared with the MIMO ARX model. Thus, the resulting multivariable multimodel, entitled MIMO ARX-Laguerre multimodel, ensures a reduction of the parametric complexity with a simple recursive vector representation. However, this reduction is conditioned by an optimal choice of the pole characterising each Laguerre base. To deal with this, we propose the identification, using the genetic algorithm, of the Laguerre poles as well as the weighting function parameters of the MIMO ARX-Laguerre multimodel. This latter allows to generate a set of linear MIMO ARX-Laguerre sub-models which are exploited to propose a robust multimodel control for uncertain nonlinear MIMO systems. In fact, for each identified sub-model, we combine the LSDP approach and the RGA (Relative Gain Array) theory to develop a local robust controller. Then, a multivariable roust multimodel control algorithm is developed where the overall control strategy by multimodel approach amounts to incorporating the same weighting functions used in the considered multivariable multimodel in order to interpolate the control laws calculated from each local robust controller. The optimisation of Laguerre poles and weighting functions parameters as well as the proposed multivariable robust multimodel control algorithm are validated on the experimental prototype Quanser Aero system.

Nomograms predicting local and distant recurrence and disease-specific mortality for R0/R1 soft tissue sarcomas of the extremities.

BackgroundPrognostic models for patients with soft tissue sarcoma (STS) of the extremities have been developed from large multi-institutional datasets with mixed results. We aimed to develop predictive nomograms for sarcoma-specific survival (SSS) and, for the first time, long-term local recurrence (LR) and distant recurrence (DR) in patients with STS of the extremities treated at our institution.Patients and methodsData from patients treated at Humanitas Cancer Center from 1997 to 2015 were analyzed. Variable selection was based on the clinical knowledge and multivariable regression splines algorithm. Perioperative treatments were always included in the model. Prognostic models were developed using Cox proportional hazards model, and model estimates were plotted in nomograms predicting SSS at 5 and 10 years and LR and DR at 2, 5, and 10 years. Model performance was estimated internally via bootstrapping, in terms of optimism-corrected discrimination (Harrell C-index) and calibration (calibration plots).ResultsData on 517 patients were analyzed. At 5 and 10 years, SSS was 68.1% [95% confidence interval (CI), 63.8–72.1] and 55.6% (50.5–60.3), respectively. LR was 79.1% (95% CI, 75.3–82.4), 71.1% (95% CI, 66.7–75.1), and 66.0% (95% CI, 60.7–70.7) at 2, 5, and 10 years, respectively, whereas DR was 65.9% (95% CI, 61.6–69.9), 57.5% (95% CI, 53.0–61.8), and 52.1% (95% CI, 47.1–56.8) at 2, 5, and 10 years, respectively. SSS nomogram included age, gender, margins, tumor size, grading, and histotype. LR and DR nomograms incorporated mostly the same variables, except for age for DR; LR nomogram did not include gender but included anatomic site. The optimism-corrected C-indexes were 0.73 and 0.72 for SSS at 5 and 10 years, respectively; 0.65, 0.64, and 0.64 for LR at 2, 5, and 10 years, respectively; and 0.68 for DR at 2, 5, and 10 years. Predicted probabilities were close to the observed ones for all outcomes.ConclusionsWe developed and validated three nomograms for STS of the extremities predicting the probability of SSS at 5 and 10 years and LR and DR at 2, 5, and 10 years. By accounting for the perioperative treatment, these models allow prediction for future patients who had no perioperative treatment, thus being useful in the clinical decision-making process.

Comparative Chemometric Manipulations of UV-Spectrophotometric Data for the Efficient Resolution and Determination of Overlapping Signals of Cyclizine and Its Impurities in Its Pharmaceutical Preparations.

Cyclizine (CYZ), a commonly used antiemetic drug, has two pharmacopeial toxic impurities, 1-methylpiperazine (MPZ) and diphenylmethanol (DPM). When CYZ parenteral formulations are administered intravenously, both impurities are poisonous, toxic, and harmful to the human body. Cyclizine was determined along with its hazardous impurities MPZ and DPM by green multivariate calibration using UV-spectroscopic data. Three multivariate algorithms were used to resolve and quantify overlapped spectral signals: principal component regression (PCR), partial least squares (PLS), and synergistic intervals partial least squares (siPLS). A concentration set containing 16 distinct combinations of CYZ, MPZ, and DPM was randomly prepared, and the absorbance values of the concentration set were determined using the 376 point-wavelength set with an interval of 0.2 nm between 200 and 275 nm. Good linear correlations were established for CYZ, MPZ, and DPM in the concentration ranges of 5.00-25.0, 0.50-2.50, and 0.50-2.50 µg/mL, respectively. The ideal spectral range and associated combinations were chosen based on the lowest root mean error of prediction (RMSEP) and correlation coefficient values (r). The siPLS approach performed better than the PCR and PLS models. The combination of four subintervals, 1, 3, 4, and 7, demonstrated the greatest effect, with RMSEP values of 0.0272, 0.0053, and 0.0315 for CYZ, MPZ, and DPM, respectively, and correlation coefficients of 0.9991, 0.9999, and 0.9997, in order. Various assessment tools were used to evaluate and measure the greenness profile of the established methods. The proposed methods were validated using internal and external validation sets. The three methods were effectively used to determine CYZ in its pure form and parenteral formulations, as well as its toxic impurities. The acquired results were compared statistically to those obtained using the reported HPLC method. Cyclizine and its toxic impurities can be determined spectrophotometrically by using the three developed chemometric models.

Beat-to-Beat Blood Pressure Estimation by Photoplethysmography and Its Interpretation

Blood pressure (BP) is among the most important vital signals. Estimation of absolute BP solely using photoplethysmography (PPG) has gained immense attention over the last years. Available works differ in terms of used features as well as classifiers and bear large differences in their results. This work aims to provide a machine learning method for absolute BP estimation, its interpretation using computational methods and its critical appraisal in face of the current literature. We used data from three different sources including 273 subjects and 259,986 single beats. We extracted multiple features from PPG signals and its derivatives. BP was estimated by xgboost regression. For interpretation we used Shapley additive values (SHAP). Absolute systolic BP estimation using a strict separation of subjects yielded a mean absolute error of and correlation of . The results markedly improve if data separation is changed (MAE: , r: ). Interpretation by means of SHAP revealed four features from PPG, its derivation and its decomposition to be most relevant. The presented approach depicts a general way to interpret multivariate prediction algorithms and reveals certain features to be valuable for absolute BP estimation. Our work underlines the considerable impact of data selection and of training/testing separation, which must be considered in detail when algorithms are to be compared. In order to make our work traceable, we have made all methods available to the public.

Non-destructive assessment of equivalent umami concentrations in salmon using hyperspectral imaging technology combined with multivariate algorithms

This study utilized equivalent umami concentrations (EUC) to characterize umami intensity in salmon with different freeze–thaw times. A rapid and non-destructive method was established to determine EUC values in salmon which is based on hyperspectral imaging (HSI) system combined with multiple characteristic variable screening methods. The established CARS-PLS model showed greater advantages in correlating the reference values of spectral data with EUC in salmon with Rc of 0.9012, Rp of 0.9009, RMSECV of 0.82, and RMSEP of 0.88. The model was employed pixel-wise to visualize the distribution of EUC with different freeze–thaw times, which demonstrated the reduction of EUC value with the increasing of freeze–thaw times. Therefore, this reseearch showed hyperspectral imaging (HSI) system combined with chemometrics possesses a substantial capability to predict and visualize the EUC of salmon, which would provide an intuitive understanding of salmon quality prediction and detection.

How heterogeneous is the dengue transmission profile in Brazil? A study in six Brazilian states.

Dengue is a vector-borne disease present in most tropical countries, infecting an average of 50 to 100 million people per year. Socioeconomic, demographic, and environmental factors directly influence the transmission cycle of the dengue virus (DENV). In Brazil, these factors vary between regions producing different profiles of dengue transmission and challenging the epidemiological surveillance of the disease. In this article, we aimed at classifying the profiles of dengue transmission in 1,823 Brazilian municipalities, covering different climates, from 2010 to 2019. Time series data of dengue cases were obtained from six states: Ceará and Maranhão in the semiarid Northeast, Minas Gerais in the countryside, Espírito Santo and Rio de Janeiro in the tropical Atlantic coast, and Paraná in the subtropical region. To describe the time series, we proposed a set of epi-features of the magnitude and duration of the dengue epidemic cycles, totaling 13 indicators. Using these epi-features as inputs, a multivariate cluster algorithm was employed to classify the municipalities according to their dengue transmission profile. Municipalities were classified into four distinct dengue transmission profiles: persistent transmission (7.8%), epidemic (21.3%), episodic/epidemic (43.2%), and episodic transmission (27.6%). Different profiles were associated with the municipality’s population size and climate. Municipalities with higher incidence and larger populations tended to be classified as persistent transmission, suggesting the existence of critical community size. This association, however, varies depending on the state, indicating the importance of other factors. The proposed classification is useful for developing more specific and precise surveillance protocols for regions with different dengue transmission profiles, as well as more precise public policies for dengue prevention.

Characterization of chromatin regulators identified prognosis and heterogeneity in hepatocellular carcinoma.

Liver carcinogenesis is a multiprocess that involves complicated interactions between genetics, epigenetics, and transcriptomic alterations. Aberrant chromatin regulator (CR) expressions, which are vital regulatory epigenetics, have been found to be associated with multiple biological processes. Nevertheless, the impression of CRs on tumor microenvironment remodeling and hepatocellular carcinoma (HCC) prognosis remains obscure. Thus, this study aimed to systematically analyze CR-related patterns and their correlation with genomic features, metabolism, cuproptosis activity, and clinicopathological features of patients with HCC in The Cancer Genome Atlas, International Cancer Genome Consortium-LIRI-JP cohort, and GSE14520 that utilized unsupervised consensus clustering. Three CR-related patterns were recognized, and the CRs phenotype-related gene signature (CRsscore) was developed using the least absolute shrinkage and selection operator-Cox regression and multivariate Cox algorithms to represent the individual CR-related pattern. Additionally, the CRsscore was an independent prognostic index that served as a fine predictor for energy metabolism and cuproptosis activity in HCC. Accordingly, describing a wide landscape of CR characteristics may assist us to illustrate the sealed association between epigenetics, energy metabolism, and cuproptosis activity. This study may discern new tumor therapeutic targets and exploit personalized therapy for patients.

A CT-based nomogram for predicting the risk of adenocarcinomas in patients with subsolid nodule according to the 2021 WHO classification

PurposeTo establish a nomogram for predicting the risk of adenocarcinomas in patients with subsolid nodules (SSNs) according to the 2021 WHO classification.MethodsA total of 656 patients who underwent SSNs resection were retrospectively enrolled. Among them, 407 patients were assigned to the derivation cohort and 249 patients were assigned to the validation cohort. Univariate and multi-variate logistic regression algorithms were utilized to identity independent risk factors of adenocarcinomas. A nomogram based on the risk factors was generated to predict the risk of adenocarcinomas. The discrimination ability of the nomogram was evaluated using the concordance index (C-index), its performance was calibrated using a calibration curve, and its clinical significance was evaluated using decision curves and clinical impact curves.ResultsLesion size, mean CT value, vascular change and lobulation were identified as independent risk factors for adenocarcinomas. The C-index of the nomogram was 0.867 (95% CI, 0.833-0.901) in derivation cohort and 0.877 (95% CI, 0.836-0.917) in validation cohort. The calibration curve showed good agreement between the predicted and actual risks. Analysis of the decision curves and clinical impact curves revealed that the nomogram had a high standardized net benefit.ConclusionsA nomogram for predicting the risk of adenocarcinomas in patients with SSNs was established in light of the 2021 WHO classification. The developed model can be adopted as a pre-operation tool to improve the surgical management of patients.

On the use of Artificial Intelligence for Condition Monitoring in Horizontal-Axis Wind Turbines

Wind power is one of the fastest-growing renewable energy sectors and is considered instrumental in the ongoing decarbonization process. However, wind turbines (WTs) present high operation and maintenance costs caused by inefficiencies and failures, leading to everincreasing attention to effective Condition Monitoring (CM) strategies. Nowadays, modern WTs are integrated with sensor networks as part of the Supervisory Control and Data Acquisition (SCADA) system for supervision purposes. CM of wind farms through predictive models based on routinely collected SCADA data is envisaged as a viable mean of improving producibility by spotting operational inefficiencies. In this paper, we introduce an unsupervised anomaly detection framework for wind turbine using SCADA data. It involves the use of a multivariate feature selection algorithm based on a novel Combined Power Predictive Score (CPPS), where the information content of combinations of variables is considered for the prediction of one or more key parameters. The framework has been tested on SCADA data collected from an off-shore wind farm, and the results showed that it successfully detects anomalies and anticipates major bearing failures by outperforming a recent deep neural approach.

The mechanisms and prediction of non-structural carbohydrates accretion and depletion after mechanical wounding in slash pine (Pinus elliottii) using near-infrared reflectance spectroscopy

BackgroundThe allocation of non-structural carbohydrates (NSCs) plays a critical role in the physiology and metabolism of tree growth and survival defense. However, little is known about the allocation of NSC after continuous mechanical wounding of pine by resin tapping during tree growth.ResultsHere, we examine the NSC allocation in plant tissues after 3 year lasting resin tapping, and also investigate the use of near-infrared reflectance (NIR) spectroscopy to quantify the NSC, starch and free sugar (e.g., sucrose, glucose, and fructose) concentrations in different plant tissues of slash pine. Spectral measurements on pine needle, branch, trunk phloem, and root were obtained before starch and free sugar concentrations were measured in the laboratory. The variation of NSC, starch and free sugars in different plant tissues after resin tapping was analyzed. Partial least squares regression was applied to calibrate prediction models, models were simulated 100 times for model performance and error estimation. More NSC, starch and free sugars were stored in winter than summer both in tapped and control trees. The position of resin tapping significantly influenced the NSCs allocation in plant tissues: more NSCs were transformed into free sugars for defensive resin synthesis close to the tapping wound rather than induced distal systemic responses. Models for predicting NSC and free sugars of plant tissues showed promising results for the whole tree for fructose (R2CV = 0.72), glucose (R2CV = 0.67), NSCs (R2CV = 0.66) and starch (R2CV = 0.58) estimates based on NIR models. Models for individual plant tissues also showed reasonable predictive ability: the best model for NSCs and starch prediction was found in root. The significance multivariate correlation algorithm for variable selection significantly reduced the number of variables. Important variables were identified, including features at 1021–1290 nm, 1480, 1748, 1941, 2020, 2123 and 2355 nm, which are highly related to NSC, starch, fructose, glucose and sucrose.ConclusionsNIR spectroscopy provided a rapid and cost-effective method to monitor NSC, starch and free sugar concentrations after continuous resin tapping. It can be used for studying the trade-off between growth and production of defensive metabolites.