Lower Prediction Error Research Articles

Multivariable Fractional Polynomials for lithium-ion batteries degradation models under dynamic conditions

Longevity and safety of lithium-ion batteries are facilitated by efficient monitoring and adjustment of the battery operating conditions. Hence, it is crucial to implement fast and accurate algorithms for State of Health (SoH) monitoring on the Battery Management System. The task is challenging due to the complexity and multitude of the factors contributing to the battery capacity degradation, especially because the different degradation processes occur at various timescales and their interactions play an important role. Data-driven methods bypass this issue by approximating the complex processes with statistical or machine learning models: they rely solely on the available cycling data, while remaining agnostic to the underlying real physical processes. This paper proposes a data-driven approach which is understudied in the context of battery degradation, despite being characterised by simplicity and ease of computation: the Multivariable Fractional Polynomial (MFP) regression. Models are trained from historical data of one exhausted cell and used to predict the SoH of other cells. The data are provided by the NASA Ames Prognostics Center of Excellence, and are characterised by varying loads which simulate dynamic operating conditions. Two hypothetical scenarios are considered: one assumes that a recent observed capacity measurement is known, the other is based only on the nominal capacity of the cell. It was shown that the degradation behaviour of the batteries under examination is influenced by their historical data, as supported by the low prediction errors achieved (root mean squared errors ranging from 1.2% to 7.22% when considering data up to the battery End of Life). Moreover, we offer a multi-factor perspective where the degree of impact of each different factor to ageing acceleration is analysed. Finally, we compare with a Long Short-Term Memory Neural Network and other works from the literature on the same dataset. We conclude that the MFP regression is effective and competitive with contemporary works, and provides several additional advantages e.g. in terms of interpretability, generalisability, and implementability.

A novel handheld FT-NIR spectroscopic approach for real-time screening of major cannabinoids content in hemp

A novel approach for rapid (15s) detection and quantification of predominant cannabinoids in hemp was developed using Fourier-transformed near-infrared spectroscopy (FT-NIR), enabling real-time and field-based applications. Hemp samples (n = 91) were obtained from certified online vendors, the OARDC Weed Lab, and a local Ohio farm. Reference data of major cannabinoids content were determined by uHPLC-MS/MS. Spectral data were collected by a miniaturized, battery-operated FT-NIR instrument, and combined with the reference data to generate partial least squares regression (PLSR) models. uHPLC-MS/MS analysis showed two samples had over 0.36% of Δ9-tetrahydrocannabinol (Δ9-THC), and 64% (32 out of 50) of online-bought hemp samples were not in compliance with their total cannabidiol (CBD) content declaration. PLSR prediction models showed excellent correlation (Rpre = 0.91–0.95) and a low standard error of prediction (SEP = 0.02–0.61%). This method could be used as an alternative to traditional methods for in-situ assessment of hemp quality.

Fuzzy logic expert system with multi-objective optimization of carbon nanotube infused nanopaint surface characteristics analysis using IRB 1410 robot

In this investigative study, the Taguchi design (L9) of experiments with TOPSIS multi-objective optimization is used for the robot nanopainting. The optimization objective is to maximize the film glue individually and minimize the surface blemishes and film thickness by varying the IRB 1410 robot painting parameters, such as the robot speed, pressure, and distance. The multi-wall carbon nanotube is infused into the paint materials using the ultrasonication process, and the mechanical properties are analysed with a comparison to regular paint materials. The virtual robot cycle time analysis is carried out for three different robot pathway patterns, namely linear, zig-zag, and circular. It is then compared with real-time experimental robot nanopaint on cold rolled close annealing steel materials. The obtained results show that surface roughness and thickness are minimized by 54% and 33.4%, respectively, using the robot nano spray coating when compared with regular spray coating. The film adhesive of robot Nano spray coating is maximized by 2.8% more than normal spray coating. A fuzzy logic expert system model is used to evaluate the surface finish characteristics of the nanopaint surface with low prediction error and with 89.2% confidence level. Heat transfer analysis of the robot nano-coated substrate is compared with that of the robot normal-coated substrate.

Construction of a nomogram model for predicting 2-year survival rate of small cell lung cancer based on more comprehensive variables

Objective: To construct a novel prognostic nomogram model based on more comprehensive variables for patients with small-cell lung cancer (SCLC). Methods: The data of 722 patients with SCLC confirmed by pathology in Affiliated Cancer Hospital of Shanxi Medical University from January 2015 to December 2018 were retrospectively analyzed [including 592 males and 130 females, aged from 23 to 82(61±9) years]. A random seed count of 133 was used to divide those patients into training set (n=422) and validation set (n=300). Kaplan-Meier was used for survival curves analysis and univariate Log-rank test was used for evaluating the influence of clinical variables on the prognosis of sclc, variables with P<0.05 in univariate analysis were included in a multivariate Cox regression model. The nomogram was constructed based on the variables which P<0.05 in multivariate analysis. Receiver operating characteristic (ROC) curve, calibration by Integrated Brier score (IBS) and clinical net benefit by decision curve analysis (DCA) were used to evaluate model discriminative power, prediction error value, and clinical net benefit, and compared with the American Joint Committee on Cancer 8th TNM. Results: Male, abnormal monocyte (MON) counts, abnormal neuron specific enolase (NSE), abnormal cytokeratin 19 fragment (Cyfra211), M1a stage, M1b stage, M1c stage, radiotherapy (RT), chemotherapy ≥4 cycles and prophylactic cranial irradiation (PCI) were prognostic factors for SCLC[HR(95%CI)=1.39(1.00-1.92), 1.29(1.02-1.63), 1.41(1.11-1.80), 2.02(1.48-2.76), 1.09(0.77-1.55), 1.44(0.94-2.22), 2.01(1.49-2.71), 0.75(0.57-0.98), 0.40(0.31-0.51)and 0.42(0.26-0.68), respectively, all P<0.05]. The area under ROC curve (AUC) of the nomogram in training set and validation set were 0.814(95%CI: 0.765-0.862)and 0.787 (95%CI: 0.725-0.849), which were higher than TNM [0.616(95%CI: 0.558-0.674) and 0.648(95%CI: 0.581-0.715)].The calibration curve showed a good correlation between the nomogram prediction and actual observation for the 2-year overall survival (OS). IBS indicted a lower prediction error rate (training set: 0.132 vs 0.169; validation set: 0.138 vs 0.169). DCA showed a wider threshold range than TNM (training set: 0.01-0.96 vs 0.01-0.85, validation set: 0.01-0.94 vs 0.01-0.86) and a greater improvement of the clinical net benefit (in training set the nomogram had a greater clinical benefit than TNM in the range of 0.19-0.96, and remained in validation set in the range of 0.19-0.94). Conclusion: The established nomogram model for predicting 2-year OS in patients with SCLC based on 8 variables, including gender, MON, NSE, Cyfra211, M stage, RT, CT cycles and PCI can be used for an more accurately prognosis prediction and reference for therapeutic regimen selection.

Temporal Convolution-Based Long-Short Term Memory Network With Attention Mechanism for Remaining Useful Life Prediction

Predictive maintenance (PdM) is useful for engineers to schedule maintenance flexibly, to operate equipment efficiently, and also to avoid unexpected downtime. Remaining useful life (RUL) prediction is critical for PdM before the need for component replacement. Data-driven approaches have attracted more attention for RUL prediction in flexible production. Compared with statistical-based and conventional machine learning approaches, deep learning-based approaches can extract critical features from raw equipment sensor data without prior knowledge from a domain expert. This study proposes a temporal convolution-based long-short term memory (TCLSTM) network with attention mechanism to extract features from equipment sensor data and to build a regression model for RUL prediction. Temporal convolutional networks (TCN) are used for feature extraction. LSTM and attention layers are used to learn temporal dependencies among the extracted features. The fully connected network consists of dense layers and is used to build the RUL prediction model. To evaluate the effectiveness of the proposed method, an empirical dataset from a semiconductor ion mill etching process was used. According to the comparison results, the proposed TCLSTM network with attention mechanism has the lowest prediction error and outperforms TCN, LSTM, and other machine learning approaches. The experimental results demonstrate the practical viability of the proposed approach.

Identification of soy sauce using high-field asymmetric waveform ion mobility spectrometry combined with machine learning

Soy sauce, an important condiment, varies greatly in the brand, geographical distribution, and production processes. We investigated the potential of volatile organic compounds (VOCs) serving as an indicator of soy sauce quality to detect three regions and two production technologies of Chinese soy sauce. An analytical method named high-field asymmetric waveform ion mobility spectrometry (FAIMS) was utilized for acquiring sample data. Wavelet packet decomposition (WPD) and principal component analysis (PCA) were used to extract the features of FAIMS data. 4 machine learning models were trained using these features, and the optimal parameters were obtained by a grid search. The scatter plots of the optimal two features we selected showed that the different regions and production technologies of soy sauce had obvious clustering trends. For the identification of different regions and production technologies, the training score, test score, and average cross-validation score of the optimal model were all 100%. Furthermore, the learning curves indicated that the optimal model obtained good performance and had low prediction errors. It was concluded that FAIMS combined with a suitable machine learning algorithm can successfully classify different regions and production technologies of Chinese soy sauce.

Multi-objective optimization of diesel engine performance and emission using grasshopper optimization algorithm

A recently invented algorithm called the grasshopper optimization algorithm (GOA) was used to predict and optimize palm oil biodiesel operated in a diesel engine. The work was conducted in three stages: (i) designing an experiment and performing the experiments, (ii) mathematical modeling, and (iii) optimization using GOA. By using regression modeling over these experimental results, the mathematical equations between the factors (biodiesel ratio (%) and load (Nm)) and the responses (BTE, BSFC, BSCO, BSNOx, BSCO2, BSHC, and Smoke) were calculated. The results showed that the factors used in the model were sufficient to explain the change in the response, and no additional factors in the mathematical models were required. The ANOVA results showed that the p-value for all the regression models were 0.000 < 0.05, which indicated their significance. Moreover, the regression models best fit the given observations with a low prediction error. The three confirmation tests also revealed satisfying results with low errors. The range of prediction error for BTE, BSFC, BSCO, BSNOx, BSCO2, BSHC, and Smoke were 0.25–3.00%, 2.55–8.20%, 4.61–11.65%, 1.71–12.20%, 1.35–3.52%, 0.02–7.75%, and 0.69–4.34%, respectively. The optimized operating conditions for the maximum engine performance and the minimum emissions was given by 50% biodiesel run at 7 Nm engine load.

Prediction of Upper Limb Action Intention Based on Long Short-Term Memory Neural Network

The use of an inertial measurement unit (IMU) to measure the motion data of the upper limb is a mature method, and the IMU has gradually become an important device for obtaining information sources to control assistive prosthetic hands. However, the control method of the assistive prosthetic hand based on the IMU often has problems with high delay. Therefore, this paper proposes a method for predicting the action intentions of upper limbs based on a long short-term memory (LSTM) neural network. First, the degree of correlation between palm movement and arm movement is compared, and the Pearson correlation coefficient is calculated. The correlation coefficients are all greater than 0.6, indicating that there is a strong correlation between palm movement and arm movement. Then, the motion state of the upper limb is divided into the acceleration state, deceleration state and rest state. The rest state of the upper limb is used as a sign to control the assistive prosthetic hand. Using the LSTM to identify the motion state of the upper limb, the accuracy rate is 99%. When predicting the action intention of the upper limb based on the angular velocity of the shoulder and forearm, the LSTM is used to predict the angular velocity of the palm, and the average prediction error of palm motion is 1.5 rad/s. Finally, the feasibility of the method is verified through experiments, in the form of holding an assistive prosthetic hand to imitate a disabled person wearing a prosthesis. The assistive prosthetic hand is used to reproduce foot actions, and the average delay time of foot action was 0.65 s, which was measured by using the method based on the LSTM neural network. However, the average delay time of the manipulator control method based on threshold analysis is 1.35 s. Our experiments show that the prediction method based on the LSTM can achieve low prediction error and delay.

Machine learning detects predictors of symptom severity and impulsivity after dialectical behavior therapy skills training group in borderline personality disorder

Only 50% of the patients with Borderline Personality Disorder (BPD) respond to psychotherapies, such as Dialectical Behavioral Therapy (DBT), this might be increased by identifying baseline predictors of clinical change. We use machine learning to detect clinical features that could predict improvement/worsening for severity and impulsivity of BPD after DBT skills training group. To predict illness severity, we analyzed data from 125 patients with BPD divided into 17 DBT psychotherapy groups, and for impulsiveness we analyzed 89 patients distributed into 12 DBT groups. All patients were evaluated at baseline using widely self-report tests; ∼70% of the sample were randomly selected and two machine learning models (lasso and Random forest [Rf]) were trained using 10-fold cross-validation and compared to predict the post-treatment response. Models’ generalization was assessed in ∼30% of the remaining sample. Relevant variables for DBT (i.e. the mindfulness ability “non-judging”, or “non-planning” impulsiveness) measured at baseline, were robust predictors of clinical change after six months of weekly DBT sessions. Using 10-fold cross-validation, the Rf model had significantly lower prediction error than lasso for the BPD severity variable, Mean Absolute Error (MAE) lasso - Rf = 1.55 (95% CI, 0.63–2.48) as well as for impulsivity, MAE lasso - Rf = 1.97 (95% CI, 0.57–3.35). According to Rf and the permutations method, 34/613 significant predictors for severity and 17/613 for impulsivity were identified. Using machine learning to identify the most important variables before starting DBT could be fundamental for personalized treatment and disease prognosis.

Novel, User-Friendly Experimental and Analysis Strategies for Fast Voltammetry: Next Generation FSCAV with Artificial Neural Networks.

Fast-scan adsorption-controlled voltammetry (FSCAV) was recently derived from fast-scan cyclic voltammetry to estimate the absolute concentrations of neurotransmitters by using the innate adsorption properties of carbon fiber microelectrodes. This technique has improved our knowledge of serotonin dynamics in vivo. However, the analysis of FSCAV data is laborious and technically challenging. First, each electrode requires post-experimental in vitro calibration. Second, current analysis methods are semi-manual and time-consuming and require a steep learning curve. Finally, the calibration methods used do not adapt to nonlinear electrode responses. In this work, we provide freely accessible computational solutions to these issues. First, we design an artificial neural network (ANN) and train it with a large data set (calibrations from 140 electrodes by six different researchers) to achieve calibration-free estimations and improve predictive error. We discuss the power of the ANN to obtain a low predictive error without electrode-specific calibrations as a function of being able to predict the sensitivity of the electrode. We use the ANN to successfully predict the absolute serotonin concentrations of real in vivo data. Finally, we create a fast and user-friendly, fully automated analysis web platform to simplify and reduce the expertise required for the postanalysis of FSCAV signals.

Long-term traffic flow prediction using multivariate SSA forecasting in SDN based networks

Accurate and real-time traffic flow forecasting plays an important role in optimizing traffic routing enabling adaptive and sophisticated applications on the network. Managing and routing enormous traffic flow with dynamic behavior is a highly challenging task. However, arriving at a precise model for traffic forecasting in a short interval of time is not trivial because of the dynamic nature of traffic flow. A novel multivariate time series framework is designed to analyze and forecast the dynamic traffic flow in SDN based networks. The proposed framework adapts the Multivariate Singular Spectrum Analysis (MSSA) forecasting model and incorporates the Randomized Singular Value Decomposition (RSVD) to improve the accuracy of flow prediction. Simulations are conducted to evaluate the effectiveness of the proposed MSSA method. The proposed method predicts the long-term traffic fluctuation from the observed traffic traces. The SDN controller is trained using the traffic traces and future traffic flows are forecasted. The performance evaluation of the proposed method predicts real-time traffic trends accurately with 2.2% MAPE, 9.44 MAE and 13.803 RMSE. The results show that the learning ability of MSSA helps to forecast future network traffic with low prediction errors.

Data prediction for cases of incorrect data in multi-node electrocardiogram monitoring

The development of a mesh topology in multi-node electrocardiogram (ECG) monitoring based on the ZigBee protocol still has limitations. When more than one active ECG node sends a data stream, there will be incorrect data or damage due to a failure of synchronization. The incorrect data will affect signal interpretation. Therefore, a mechanism is needed to correct or predict the damaged data. In this study, the method of expectation-maximization (EM) and regression imputation (RI) was proposed to overcome these problems. Real data from previous studies are the main modalities used in this study. The ECG signal data that has been predicted is then compared with the actual ECG data stored in the main controller memory. Root mean square error (RMSE) is calculated to measure system performance. The simulation was performed on 13 ECG waves, each of them has 1000 samples. The simulation results show that the EM method has a lower predictive error value than the RI method. The average RMSE for the EM and RI methods is 4.77 and 6.63, respectively. The proposed method is expected to be used in the case of multi-node ECG monitoring, especially in the ZigBee application to minimize errors.

Gated recurrent unit least-squares generative adversarial network for battery cycle life prediction

One of the main concerns of battery management systems is predicting the degradation of lithium-ion batteries, which remaining useful life prediction is an essential tool for prognostic and health management of batteries. In this study, we develop a novel prognostic architecture that is based on a least-squares generative adversarial network with the gated recurrent unit as the generator and multi-layer perceptron as the discriminator and use it to predict the Lithium-ion batteries’ remaining useful life. The proposed method aims to learn the probability distribution of future values in an adversarial training fashion. This generative adversarial network gives more penalties to large errors and addresses the vanishing gradient problem during training. As a result, the predicted values will get closer to the actual data. Furthermore, to obtain high prediction accuracy, time-domain features are evaluated using statistical formulas. The most important features are then selected using the random forest algorithm and fed to the network as a multivariate input set. The performances of the proposed method are tested using a battery degradation dataset from the data repository of Prognostics Center of Excellence at NASA. Furthermore, experimental data from lithium-ion cells at different current rates are conducted for evaluation and verification. The obtained outcomes demonstrate that the designed model achieves the low prediction error of 2.63% and maximum absolute error of 0.02.

Novel cross LSTM for predicting the changes of complementary pelvic angles between standing and sitting.

Sagittal spino-pelvic balance has been increasingly emphasized in hip surgery. The conversion between standing and sitting, characterized by complementary pelvic angles (pelvic tilt, pt and sacral slope, ss), involves a congruent sagittal spino-pelvic relationship. Hence, the changes of complementary pelvic angles pt, ss between standing and sitting could reflect the mechanism of sagittal spino-pelvic balance, and should be analyzed in evidence-based hip surgery planning. To this end, we propose a novel cross LSTM (C-LSTM) framework embedding the conversion between standing and sitting by cross-mapping, to predict the changes of complementary pelvic pt, ss between standing and sitting. Furthermore, to introduce the prior knowledge of the invariance of pelvic incidence, pi, two dual C-LSTMs are integrated to construct a much more powerful Fused C-LSTM. We have conducted extensive experiments on the sagittal standing-sitting dataset for the comprehensive evaluation of the proposed framework. Even in a small samples, Fused C-LSTM can achieve low prediction errors and high correlation between predicted and actual values. Notably, just based on static standing or sitting X-ray, Fused C-LSTM can obtain the change of complementary pt, ss between standing and sitting to assist in formulating a surgical hip plan that conforms to the sagittal spino-pelvic balance.

Influence of the Intraday Electricity Market Structure on the Degradation of Li‐Ion Batteries Used to Firm Photovoltaic Production

This article considers the introduction of Li‐ion batteries in photovoltaic power plants to firm their energy production and analyzes the dependence of their degradation on the structure of the electricity market where the power production is traded. The operation of the batteries is decided as a result of successive optimization problems that benefit from the use of deep‐learning‐based irradiance forecasting tools with low prediction error, which allows the batteries to keep a small size. In addition, state‐of‐the‐art battery aging models are handled to derive a realistic lifetime prognosis. The simulation results obtained by using real data from three European locations, which have different irradiance patterns and market structures, show how the proposed control strategy makes it possible to decouple the saturation rate of the batteries from the climatic conditions of the plant location. Furthermore, regarding the market structure, the results show that the shorter the energy block and the closer the lead time, the lower is the degradation of the batteries.

Nonlinear spatial and temporal decomposition provides insight for climate change effects on sub-Arctic herbivore populations

Global temperatures are increasing, affecting timing and availability of vegetation along with relationships between plants and their consumers. We examined the effect of population density, herd body condition in the previous year, elevation, plant productivity and phenology, snow, and winter onset on juvenile body mass in 63 semi-domesticated populations of Rangifer tarandus throughout Norway using spatiotemporal generalized additive models (GAMs) and varying coefficient models (VCMs). Optimal climate windows were calculated at both the regional and national level using a novel nonlinear climate window algorithm optimized for prediction. Spatial and temporal variation in effects of population and environmental predictors were considered using a model including covariates decomposed into spatial, temporal, and residual components. The performance of this decomposed model was compared to spatiotemporal GAMs and VCMs. The decomposed model provided the best fit and lowest prediction errors. A positive effect of herd body condition in the previous year explained most of the deviance in calf body mass, followed by a more complex effect of population density. A negative effect of timing of spring and positive effect of winter onset on juvenile body mass suggested that a snow free season was positive for juvenile body mass growth. Our findings suggest early spring onset and later winter permanent snow cover as reinforcers of early-life conditions which support more robust reindeer populations. Our methodological improvements for climate window analyses and effect size measures for decomposed variables provide important contributions to account for, measure, and interpret nonlinear relationships between climate and animal populations at large scales.

Validating a spatio-temporal model of observed neighborhood physical disorder

This study tested spatio-temporal model prediction accuracy and concurrent validity of observed neighborhood physical disorder collected from virtual audits of Google Street View streetscapes. We predicted physical disorder from spatio-temporal regression Kriging models based on measures at three dates per each of 256 streestscapes (n = 768 data points) across an urban area. We assessed model internal validity through cross validation and external validity through Pearson correlations with respondent-reported perceptions of physical disorder from a breast cancer survivor cohort. We compared validity among full models (both large- and small-scale spatio-temporal trends) versus large-scale only. Full models yielded lower prediction error compared to large-scale only models. Physical disorder predictions were lagged at uniform distances and dates away from the respondent-reported perceptions of physical disorder. Correlations between perceived and observed physical disorder predicted from the full model were higher compared to that of the large-scale only model, but only at locations and times closest to the respondent's exact residential address and questionnaire date. A spatio-temporal Kriging model of observed physical disorder is valid.

Detecting Intra-Field Variation in Rice Yield With Unmanned Aerial Vehicle Imagery and Deep Learning.

Unmanned aerial vehicles (UAVs) equipped with multispectral sensors offer high spatial and temporal resolution imagery for monitoring crop stress at early stages of development. Analysis of UAV-derived data with advanced machine learning models could improve real-time management in agricultural systems, but guidance for this integration is currently limited. Here we compare two deep learning-based strategies for early warning detection of crop stress, using multitemporal imagery throughout the growing season to predict field-scale yield in irrigated rice in eastern Arkansas. Both deep learning strategies showed improvements upon traditional statistical learning approaches including linear regression and gradient boosted decision trees. First, we explicitly accounted for variation across developmental stages using a 3D convolutional neural network (CNN) architecture that captures both spatial and temporal dimensions of UAV images from multiple time points throughout one growing season. 3D-CNNs achieved low prediction error on the test set, with a Root Mean Squared Error (RMSE) of 8.8% of the mean yield. For the second strategy, a 2D-CNN, we considered only spatial relationships among pixels for image features acquired during a single flyover. 2D-CNNs trained on images from a single day were most accurate when images were taken during booting stage or later, with RMSE ranging from 7.4 to 8.2% of the mean yield. A primary benefit of convolutional autoencoder-like models (based on analyses of prediction maps and feature importance) is the spatial denoising effect that corrects yield predictions for individual pixels based on the values of vegetation index and thermal features for nearby pixels. Our results highlight the promise of convolutional autoencoders for UAV-based yield prediction in rice.

Deep Learning-Based Destination Prediction Scheme by Trajectory Prediction Framework

Destination prediction is an important task in vehicular ad hoc networks (VANETs), which benefits to resolve traffic congestion. The traditional destination prediction methods are mainly based on statistical learning, which is difficult to fit the movement pattern of vehicles and predict destination of a new query trajectory. In recent years, deep learning attracted more and more attentions since it has a better ability to adaptively extract features from data and fit complex models. The exiting deep learning-based destination prediction methods mainly extract features from past trajectory and fail to exploit future trajectory because the future locations are unknown, which results in a large prediction error. In this paper, we propose a novel deep learning-based destination prediction model that can predict the future trajectory and make full use of the information about it. Firstly, we exploit a trajectory prediction method to predict future locations, which makes the process of destination prediction continuous. Secondly, we design a branch network to determine which future location is the destination of the trajectory. We also evaluate our model with a real-world dataset. Experimental results show that our model can predict the coordinates of destination with a low prediction error.

Near-infrared spectroscopy with chemometrics for identification and quantification of adulteration in high-quality stingless bee honey

This study presents a simple, rapid, and non-destructive approach of combining near-infrared spectroscopy (NIRS) with chemometrics for the evaluation of adulteration levels in stingless bee honey (SBH). Three high-quality SBH samples were directly obtained from the northern part of Malaysia and then adulterated with water (W) and apple cider vinegar (AC) at an adulteration range of 4.76%–50%. The NIRS analysis was performed at the spectral region of 700–1100 ​nm. The chemometric tools used include principal component analysis (PCA), hierarchical cluster analysis (HCA), principal component analysis-linear discriminant analysis (PCA-LDA), and partial least squares regression (PLSR). Using the first three principal components (PCs) with a total of 97.95% of explained variance, a complete distinction between adulterants, pure, and adulterated samples was achieved. PCA-LDA model with 100% classification accuracy in prediction was able to discriminate each SBH adulterated with W and AC. A general PLSR model to quantify the level of adulteration was developed. The best prediction model used 7 factors with a high correlation coefficient ‘R’ and low root mean square error of prediction ‘RMSEP’ (RP ​= ​0.995 and RMSEP ​= ​1.350%). These results confirm the combined ability of NIRS at region 700–1100 ​nm and chemometrics to effectively discriminate and quantify adulterated SBHs.