Glass Dataset Research Articles

Behavioral analysis of electricity consumption characteristics for customer groups using the k-means algorithm

In the fierce competition of the electricity market, how to consolidate and develop customers is particularly important. Aiming to analyze the electricity consumption characteristics of customer groups, this paper used a k-means algorithm and optimized it. The number of clusters was determined by the Davies-Bouldin index (DBI). An improved Harris Hawks optimization (IHHO) algorithm was designed to realize the initial cluster center selection. Based on data such as electricity purchase and average electricity price, electricity customer groups were clustered using the IHHO-k-means algorithm. The IHHO-k-means algorithm achieved the best clustering effect on Iris, Wine, and Glass datasets compared with the traditional k-means and PSO-k-means algorithms. Taking Iris as an example, the optimal value of the IHHO-k-means algorithm was 96.538, with an accuracy rate of 0.932, precision and recall rates of 0.941 and 0.793, respectively, an F-measure of 0.861, and an area under the curve (AUC) value of 0.851. In the customer dataset, the number of clusters determined by DBI was 4. The power customers were divided into four groups with different characteristics of electricity consumption, and their electricity consumption behaviors were analyzed. The results prove the reliability of the IHHO-k-means algorithm in analyzing electricity consumption characteristics of customer groups, and it can be applied in practice.

Dissecting the Characteristics and Driver Factors on Global Water Use Efficiency Using GLASS Data Sets

AbstractEcosystem water use efficiency (WUE) is a crucial parameter for understanding the interaction between carbon and water cycles. However, the spatio–temporal evolution and drivers of WUE remain unclear. This study utilized global annual scale global land surface satellite gross primary productivity and evapotranspiration data from 1982 to 2018 to estimate WUE and analyze its spatio–temporal characteristics. Additionally, the study investigated the response of WUE changes to five environmental factors (precipitation [PRE], soil moisture, temperature [TEM], palmer drought severity index, and vapor pressure deficit [VPD]) on WUE changes using partial correlation and structural equation modeling. The results suggested that the global annual WUE increased markedly over the study period, at an average rate of 0.0016 gC m−2 mm−1 H2O year−1. In contrast to the existing knowledge on the drivers of WUE change, climate change was found to have a larger contribution to WUE changes at the global and regional scales, especially in terms of TEM and VPD. A positive correlation between TEM and WUE was observed, but extreme TEM could lead to a decrease in WUE. VPD had the most significant direct effect on WUE, and its negative effect offset the positive influence of TEM especially in hyper‐arid, semi‐arid, and arid regions. These findings offer new insights into the impact of VPD and global warming on WUE.

Deep learning-based prediction of saturation magnetic flux density in Fe-based metallic glasses via transfer learning

Transfer learning (TL) is a widely recognized machine learning model that leverages domain knowledge from one or more predictive tasks to enhance performance in a related task. This study demonstrated the improved prediction performance of saturation magnetic flux density (Bs) of bulk metallic glasses (BMGs) with deep learning through the utilization of TL, which enabled the extension of compositional knowledge to overcome the limitations posed by scarce data. Specifically, TL was demonstrated by training a convolutional autoencoder network on full BMG datasets and transferring the acquired compositional knowledge to scarce iron (Fe)-based metallic glass datasets. Through this transfer, the broad compositional knowledge derived from the full BMG dataset is embedded in data representations, allowing its domain knowledge to be shared with smaller datasets. Here, a convolutional neural network was used as a predictor, with the prediction performance of Bs evaluated by the determination coefficient (R2 score) and root-mean-squared error (RMSE). The predictor, having received knowledge about the broad composition through TL, exhibited enhanced predictive performance of Bs on unseen BMG datasets while maintaining its prediction capability for Fe-based metallic glass alloys with R2 = 0.939 and RMSE = 0.084 T. This performance surpassed the performance of directly transferred knowledge on scarce data. These findings underscore the benefits of TL in predictive tasks with limited datasets and highlight the potential for applying this method to the development of new materials, where well-trained models can be repurposed for unlabeled material datasets with desired properties.

BIOM-17. WHOLE GAIN OF CHROMOSOME 19, NOT CO-GAIN OF CHROMOSOMES 19 AND 20, IS A FAVORABLE PROGNOSTIC BIOMARKER IN IDH-WILDTYPE GLIOBLASTOMA

Abstract BACKGROUND IDH-wildtype glioblastoma carries an abysmal prognosis, but long-term survivors exist. Previous studies have suggested that co-gain of chromosomes 19 and 20 (+19/+20) is a favorable prognostic marker. We sought to validate this claim and investigate whether gain of chromosome 19 (+19) is a superior prognostic indicator. METHODS We collected somatic copy number alteration data for primary glioblastoma tumors from TCGA (N = 356) and paired primary (N = 169) and recurrent (N = 167) glioblastoma tumors from GLASS. We assessed the frequency of these events and conducted survival analyses with univariate and multivariate Cox proportional hazard models to examine the association between +19/+20 and +19 and patient outcomes. RESULTS Among primary glioblastomas, the frequency of +19/+20 (20.8% vs. 23.7%) and +19 (34.6% vs. 36.7%) did not differ between TCGA and GLASS. However, the presence of +19/+20 was significantly reduced in GLASS recurrent tumors compared to primary tumors (14.3%, p = 0.04), while +19 without +19/+20 was unchanged (13.8%), indicating that +19/+20 is selected against in recurrent glioblastoma. Tumors harboring +19/+20 did not exhibit improved overall survival (OS) or progression-free survival (PFS) in TCGA (p = 0.21 and p = 0.14, respectively), nor did they demonstrate improved OS in GLASS (p = 0.32). Conversely, tumors with +19 showed significantly better OS and PFS in TCGA (HR = 1.4, p< 0.01; HR = 1.5, p< 0.01, respectively) and improved OS in GLASS (HR = 1.6, p< 0.01). In multivariate survival analyses, +19 remained a prognostic factor independent of +19/+20, patient sex, age, MGMT promoter methylation status, and KPS in TCGA (p< 0.05) and +19/+20 and age in GLASS (p = 0.01). CONCLUSION Although previous studies have suggested that +19/+20 is a positive prognostic biomarker in glioblastoma, our findings indicate that this assertion should be reevaluated. Our results demonstrate that +19 is associated with favorable outcomes in the TCGA and GLASS datasets, whereas +19/+20, while possibly selected against in recurrent glioblastomas, is not.

AT-densenet with salp swarm optimization for outlier prediction

In the field of data mining, outlier prediction detects objects that behave differently from normal objects. The conventional density and distance based outlier detection cannot identify the outliers present generally in the stream data. Transfer learning applies the knowledge of Artificial Intelligence to the task of outlier prediction. It makes training faster than earlier stages of the deep learning model. However, the correlations between spatial points are not detected with the Transfer learning based approach. It can be resolved with attention based approach, which focuses on the most informative data points. Hence, this work introduces Transfer learning with an attention mechanism based deep learning model for outlier detection. The attention transfer-DenseNet model is utilized for the outlier prediction. The attention mechanism highlights outlier information in feature maps and enhances the prediction performance. Further, the attention transfer-DenseNet weights are optimized by the metaheuristic algorithm salp swarm optimization algorithm. The performance of the proposed outlier prediction is compared over the conventional models and attained better mean average precision of 0.98 (Glass dataset) and 0.986 (Abalone dataset), respectively.

Personal Protective Equipment Detection: A Deep-Learning-Based Sustainable Approach

Personal protective equipment (PPE) can increase the safety of the worker for sure by reducing the probability and severity of injury or fatal incidents at construction, chemical, and hazardous sites. PPE is widely required to offer a satisfiable safety level not only for protection against the accidents at the aforementioned sites but also for chemical hazards. However, for several reasons or negligence, workers may not commit to and comply with the regulations of wearing the equipment, occasionally. Since manual monitoring is laborious and erroneous, the situation demands the development of intelligent monitoring systems to offer the automated real-time and accurate detection of PPE compliance. As a solution, in this study, Deep Learning and Computer Vision are investigated to offer near real-time and accurate PPE detection. The four colored hardhats, vest, safety glass (CHVG) dataset was utilized to train and evaluate the performance of the proposed model. It is noteworthy that the solution can detect eight variate classes of the PPE, namely red, blue, white, yellow helmets, head, person, vest, and glass. A two-stage detector based on the Fast-Region-based Convolutional Neural Network (RCNN) was trained on 1699 annotated images. The proposed model accomplished an acceptable mean average precision (mAP) of 96% in contrast to the state-of-the-art studies in literature. The proposed study is a potential contribution towards the avoidance and prevention of fatal/non-fatal industrial incidents by means of PPE detection in real-time.

An improved K-Means clustering based on differential evolution

Randomly selecting the initial cluster center point in the K-means clustering algorithm can lead to the issue of local optima. The paper introduces an enhanced differential evolution algorithm for optimizing K-means clustering. Experimental results demonstrate the superiority of the improved K-means clustering, based on the proposed differential evolution, over other comparison algorithms. Notably, the optimized version exhibits exceptional performance in clustering Iris, Wine, and Glass datasets. The initial clustering center can improve the clustering accuracy in the later stage. Therefore, the proposed algorithm is more superior.

Antlion Optimizer Algorithm Modification for Initial Centroid Determination in K-means Algorithm

Clustering is a grouping of data used in data mining processing. K-means is one of the popular clustering algorithms, is easy to use, and is fast in clustering data. The K-means method groups the data based on k distances and randomly determines the initial centroid as a reference for processing. Careless selection of centroids can result in poor clustering processes and local optima. One of the improvements in determining the initial centroid on the k-means method is to use the optimization method to determine the initial centroid. The modified Antlion Optimizer (ALO) method is used to improve poor clustering in the initial centroid determination and as an alternative to determining the initial centroid in the k-means method for better clustering results. The results of the research on the use of the proposed method for determining the initial centroid provide an increase in clustering compared to the usual k-means and k-means++ methods. This is evidenced by the evaluation of the sum of intragroup distance (SICD) with UCI datasets, namely iris, wine, glass, ecoli, and cancer, in each method, the best SICD value was obtained in the proposed method. Then measuring the best SICD value for each method and dataset is measured by providing a ranking proving that the proposed method on the iris, wine, and cancer datasets gets the first rank, and on the ecoli and glass datasets the proposed method and the k-means++ method both get the first rank. From the average ranking value, the proposed method is ranked first, which provides evidence that the proposed method can improve the clustering results and can be an alternative method for determining the initial center of a cluster using the k-means method.

Cluster analysis for the selection of potential discriminatory variables and the identification of subgroups in archaeometry

In this article, three variable selection methods based on Gaussian mixture models were compared to find a subset of variables that provided the “best” clustering. The use of an appropriate transformation for compositional data, whose geometric space is the Simplex, is emphasized. The comparison revealed the ability of the models to cluster data in multiple phases, showing to be more convenient to select the relevant variables than to perform an analysis based on 2D plots or by simultaneously including all the available variables in a multivariate analysis. Once the informative variables for the clustering were obtained, we used a method called Databionic Swarm (DBS). This method uses unsupervised machine learning, taking advantage of emergence and swarm intelligence applied to find natural chemical groups in the input data space. DBS can visualize high-dimensional distances in the projection through a 3D topographic map with hypsometric tints. The results were compared in terms of accuracy, both in the selection of the variables and in the classification, using a supervised accuracy index for clustering and two unsupervised indexes (the Heatmap and the Silhouette plot). The concepts and methods were illustrated by applying them to two published archaeological glass data sets. The first set consisted of 245 Romano-British glass vessels and the second set of 180 glass vessels from the 15th-17th century in Antwerp. In these applications, it was found that the methods for the selection of variables increased the accuracy of the classification compared to traditional methods.

On cobalt zinc boro sodium fluoride glasses doped with Y2O3: Synthesis, artificial intelligence density prediction and dielectric spectroscopy

Artificial intelligence density prediction and dielectric spectroscopy of five glass samples with chemical compositions 64.9B2O3-25NaF-10ZnO-0.1CoO-xY2O3: x = 0.0 (Y0.0), 0.2 (Y0.2), 0.4 (Y0.4), 0.8 (Y0.8), and 1.0 (Y1.0) mol% have been investigated. Fabrication process has been achieved by melting quenching method at melting temperature 1050–1100 °C. Density of the Y0.0-Y1.0 samples experimentally was measured and varied from 2.945 to 3.016 g/cm3. The chemical composition of the prepared glasses successfully mapped with various machine learning (ML) algorithms to predict densities using 10,000 glass data set. The density prediction done by various machine learning (ML) techniques like the logistic regression, linear regression, and decision tree (DT), random forest regression (RFR), and artificial neural network (ANN). A decision tree algorithm being fitted to the training set, estimating the test outcome utilizing the confusion matrix. The accuracy of 78.261 as well as a classification error equal 0.392 obtained for our data set which was best values to predict density for various ML techniques. The logistic regression also predicting successfully the glass data with R2 0.850 compare to linear regression. The random forest regression dominating in density prediction in current study with maximum R2 values 0.987. In addition, the dielectric spectroscopy of the prepared Y0.0-Y1.0 glasses is examined through various frequencies at room temperature (RT) to investigate the structure changes by yttrium doping on dielectric spectroscopy items. Results revealed that the insertion of yttrium ion causes an enhancement in dielectric constant and the dependency of dielectric loss (tan δ) on applied frequency has relaxation peaks at around 700 and 7000 Hz.

Exploring a general convolutional neural network-based prediction model for critical casting diameter of metallic glasses

Metallic glasses (MGs) as emerging amorphous materials have attracted considerable interest due to their excellent mechanical, physical, and chemical properties. However, the poor glass-forming ability (GFA) of MGs makes them difficult to produce large specimen sizes required by engineering applications. Traditional design on MGs mainly relies on empirical rules, thus being limited in the accuracy and generality, in turn leading to slow development of MGs. Motivated by the issue, we hope to utilize the powerful learning capability of deep learning to mine the relationship between the metallic glass structure and the critical casting diameter (Dmax) such that develop a universal and accurate tool to aid the experimental investigation. Based on a reliable metallic glass dataset of 1121 unique alloys collected from literature, we introduce a periodic table representation (PTR) strategy to characterize the MG structure, which only needs the information of alloy composition such that can avoid feature engineering involving domain knowledge. Based on the image representation, we accordingly construct a convolutional neural network to effectively extract the structure feature from PTR, and then follow a fully connected neural network to realize the Dmax prediction. In addition, two data augmentation strategies are introduced to address the dependence of deep learning on big data, through which the model performance is indeed improved. In particular, the pairwise difference regression (PADRE) strategy exhibits better performance than the Mixup augmentation way, as PADRE targets the difference between sample pairs such that can to some extent drop the systematic error of experimental determination. The principal component analysis (PCA) analysis further confirms the effectiveness of the PTR representation and the powerful capacity of CNN to extract structural information from PTR. Benefited from the technical advantages, our deep learning model achieves satisfactory performance with R2 of 0.822 for unseen samples in the independent test set, which outperforms two competitive models using traditional machine learning algorithms coupled with hand-selection features, further demonstrating the advantage of our model in the universality and accuracy.

Impacts of Extreme Climates on Vegetation at Middle-to-High Latitudes in Asia

In this study, we investigated the synchronous responses of vegetation to extreme temperatures and/or precipitation at middle-to-high latitudes in Asia using semi-monthly observations of the GIMMS and GLASS leaf area index (LAI) from 1982 to 2016. The extreme vegetation and climate states were specified using standard anomalies of the annual cycle with removed variables. The results show that the area with the maximum or minimum LAI increased or decreased in correspondence with global warming. Both the GIMMS and GLASS LAI mostly reached their maximum in spring and autumn. The GIMMS LAI mostly reached its minimum in summer, while the GLASS LAI mostly reached its minimum in late spring or late summer. The GIMMS and GLASS datasets were generally consistent regarding the vegetation responses to extreme temperatures and precipitation, especially in the areas covered by trees. Extreme cold and/or wet conditions inhibited forest growth in the area south of 60 °N, particularly from October to November. Extreme hot and/or dry conditions promoted forest growth, particularly in the central and northern parts of Siberia from August to September. However, in some arid areas of Central Asia and the Mongolian Highlands, which are mostly covered by sparse vegetation and grasses, low temperature extremes and/or strong precipitation promoted vegetation growth, while high temperature extremes and/or low precipitation had adverse effects on vegetation growth. This was more apparent in the GIMMS LAI than it was in the GLASS LAI, since the GIMMS dataset supplied more values representing extreme states of vegetation. The compound extreme of hot-and-dry and cold-and-wet climates were more frequent than the combination of cold and dry climates and hot-and-wet climates were. The overall positive response of the vegetation was superior to the negative response. The results of this study suggest that a continuous increase in vegetation density and coverage will occur over the boreal region in the future if the warming trend persists. The consequent climate feedback in this area on the regional and global scales should be afforded more attention.

Assessing the Fitness of Satellite Albedo Products for Monitoring Snow Albedo Trends

Accurate monitoring of albedo trends over snow is essential to evaluate the consequences of the global snow cover retreat on Earth’s energy budget. Satellite observations provide the best way to monitor these trends globally, but their uncertainty increases over snow. Besides, different products sometimes show diverging trends. A better assessment of the fitness of satellite products for monitoring snow albedo trends is needed. We analyze the consistency of black-sky albedo estimates from global long-term products over snow: AVHRR-based (CLARA-A2.1, GLASS-v4.2), MODIS-based (MCD43C3-v6.1/v6, GLASS-v4.2), MISR-based (MIL3MLSN-v4) and multi-sensor (C3S-v1/v2). We use MCD43C3-6.1 as the reference based on a previous comparison against in-situ measurements. CLARA-A2.1 is the one most consistent with MCD43C3, but has a low coverage in high latitudes and an artificial albedo decrease since 2015. The study shows the limitations of MIL3MLSN, GLASS, and C3S multi-sensor products over snow. MIL3MLSN has a too-low coverage of albedo over snow. GLASS-AVHRR overestimates albedo in regions with seasonal snow due to delayed snowmelt and underestimates it in permanently snow-covered regions. GLASS-MODIS is more consistent with MCD43C3 at mid-latitudes, but also underestimates albedo in regions with permanent snow and has an increase of missing values after 2011. Both GLASS datasets are temporally inconsistent with the other products. Despite the improvements from v1 to v2, C3S-v2 has the largest negative bias over snow and discontinuities in the transitions between sensors. The study evidences the difficulties of AVHRR products to provide stable snow albedo estimates in polar regions, particularly before 2000.

Neural network with NewSigmoid activation function

We increase the power of the Artificial Neural Networks with the help of the Activation Function (AF). The tansig and logsig are widely used AF. But there is still requires some improvement in the AF. So, in this paper, we have proposed a NewSigmoid AF in the neural network. NewSigmoid is also as powerful as tansig and logsig. In multiple cases, the NewSigmoid function gives a better or equivalent performance as compared with both these AF. Like these AF, NewSigmoid is also a smooth S-shape, bounded, continuously differentiable, and zero-centered function. Therefore the NewSigmoid is also suitable for solving non-linear problems. We have tested this AF on iris, cancer, glass, chemical, bodyfat, wine, and ovarian datasets. We use Scaled Conjugate Gradient (SCG), Levenberg-Marquardt (LM), and Bayesian Regularization (BR) algorithms during the optimization of the neural network. Maximum 100% accuracy in the iris dataset while using LM, and BR; 99.9% accuracy in the cancer dataset using BR; 100% accuracy in the glass dataset using BR; 100% accuracy in the chemical and bodyfat dataset using SCG, LM, and BR; 100% accuracy in the wine dataset using LM, and BR; and 99.1% accuracy in the ovarian dataset using BR has been found while working with multilayer neural networks. The NewSigmoid also achieves 100% training and validation accuracy on the mathework-cap image dataset using SCG.

Glass Defect Detection via Multi-Scale Feature Fusion

Glass defect detection is significant in glass industry. However, most of the existing methods for glass defect detection currently still rely on manual screening with high-cost and poor-efficiency. To address this issue, we propose a glass defect detection method using multi-scale feature fusion strategy. Specifically, we first propose an algorithm based on pix2pix to realize the edge extraction. Then, we propose a glass defect detector based on both local and global features. Comprehensive experiments are conducted on collected glass dataset from factories. The experimental results demonstrate that our method outperforms conventional methods including the traditional image processing and Holistically-Nested Edge Detection (HED), with the precision rate (PR) up to 97%, the false precision rate (FP) below 2% and the total accuracy rate (ACC) of glass defect detection up to 98%.

Quantum circuit design search

This article explores search strategies for the design of parameterized quantum circuits. We propose several optimization approaches including random search plus survival of the fittest, reinforcement learning both with classical and hybrid quantum classical controllers, and Bayesian optimization as decision makers to design a quantum circuit in an automated way for a specific task such as multi-labeled classification over a dataset. We introduce nontrivial circuit architectures that are arduous to be hand-designed and efficient in terms of trainability. In addition, we introduce reuploading of initial data into quantum circuits as an option to find more general designs. We numerically show that some of the suggested architectures for the Iris dataset accomplish better results compared to the established parameterized quantum circuit designs in the literature. In addition, we investigate the trainability of these structures on the unseen dataset Glass. We report meaningful advantages over the benchmarks for the classification of the Glass dataset which supports the fact that the suggested designs are inherently more trainable.

An improved Kohonen self-organizing map clustering algorithm for high-dimensional data sets

<p>Manipulating high-dimensional data is a major research challenge in the field of computer science in recent years. To classify this data, a lot of clustering algorithms have already been proposed. Kohonen self-organizing map (KSOM) is one of them. However, this algorithm has some drawbacks like overlapping clusters and non-linear separability problems. Therefore, in this paper, we propose an improved KSOM (I-KSOM) to reduce the problems that measures distances among objects using EISEN Cosine correlation formula. So far as we know, no previous work has used EISEN Cosine correlation distance measurements to classify high-dimensional data sets. To the robustness of the proposed KSOM, we carry out the experiments on several popular datasets like Iris, Seeds, Glass, Vertebral column, and Wisconsin breast cancer data sets. Our proposed algorithm shows better result compared to the existing original KSOM and another modified KSOM in terms of predictive performance with topographic and quantization error.</p>

Density of fluoride glasses through artificial intelligence techniques

Artificial Intelligence techniques have been employed for the first time to train a dataset of 1258 distinct fluoride glasses collected from published literature to predict the density of novel oxy-fluoro glasses based on their chemical composition and ionic radii. The glass dataset was split based on the linear and non-linear variation to predict the glass density using various AI models like gradient descent, random forest regression and artificial neural networks. High concentration of boron in glass specimens resulted in scattering of datapoints in packing factor relation and density prediction. The random forest regression model fit the combined glass dataset with the highest R2 of 0.980. In case of boron-rich glasses, their non-linear behavior restricted the R2 for ANNs to 0.792 as optimum with the tanh activation function.

A Comparison between Seven Heuristic Methods to Estimate the Number of Hidden Layer Neurons in a Multi Layer Perceptron Neural Network

Multilayer Perceptron Neural Network (MLPNNs) constructs of input, at least one hidden and output layer. Number of the neurons in the hidden layer affects the NNs performance. It also consider difficult task to overcome. This research, aims to exanimate the performance of seven heuristic methods that have been used to estimate the neurons numbers in the hidden layer. The effectiveness of these methods was verified using a six of benchmark datasets. The number of hidden layer neurons that selected by each heuristic method for every data set was used to train the MLP. The results demonstrate that the number of hidden neurons selected by each method provides different accuracy and stability compared with other methods. The number of neurons selected by Hush method for ine data set was 26 neurons. It’s achieved the best accuracy with 99.90%and lowest accuracy achieved by Sheela method with 67.51% using 4 neurons. Using 22 neurons with 97.97% accuracy Ke, J method received the best result for Ionosphere data set. While the lowest accuracy was 96.95% with 5 neurons achieved by Kayama method.For Iris data set with 8 neurons achieved 97.19 as best accuracy achieved by Hush method. For the same data set the lowest results were 92.33 % using 3 neurons obtained by using Kayama method. For WBC data set 96.40% the best accuracy achieved using Sheela and Kaastra methods using 4 and 7neurons, while Kanellopoulos method achieved the lowest accuracy 94.18% with 7neurons. For Glass dataset, 87.15% was the best obtained accuracy using 18 neurons Hush method and using Wang method 82.27 % with 6 neurons was the lowest accuracy. Finally for PID 75.31% accuracy achieved by Kayama method with 3 neurons, where Kanellopoulos method obtained 72.17% through using 24 neurons.

Evaluation of Evapotranspiration Models Using Different LAI and Meteorological Forcing Data from 1982 to 2017

We evaluated the performance of three global evapotranspiration (ET) models at local, regional, and global scales using the multiple sets of leaf area index (LAI) and meteorological data from 1982 to 2017 and investigated the uncertainty in ET simulations from the model structure and forcing data. The three ET models were the Simple Terrestrial Hydrosphere model (SiTH) developed by our team, the Priestley–Taylor Jet Propulsion Laboratory model (PT-JPL), and the MODerate Resolution Imaging Spectroradiometer (MODIS) ET algorithm (MOD16). Comparing the observed with simulated monthly ET by the three models over 43 Fluxnet sites, we found that SiTH overestimated ET for forests with mean slope from 1.25 to 1.67, but it performed better than the other two models over short vegetation. MOD16 and PT-JPL models simulated well for forests but poorly in dryland biomes (slope = 0.25~0.55; R2 = 0.02~0.46). At the catchment scale, all models performed well, except for some tropical and high latitudinal catchments, with NSE values lower than 0 and RMSE and MAE values far beyond their mean values. At the global scale, SiTH highly overestimated ET in tropics, while PT-JPL slightly underestimated ET between 30°N and 60°N and MOD16 underestimated ET between 15°S and 30°S. Generally, the PT-JPL provided the better performance than SiTH and MOD16 models. This study also revealed that the estimated ET by SiTH and especially PT-JPL model were influenced by the uncertainty in meteorological data, and the estimated ET was performed better using MERRA-2 datasets for PT-JPL and using ERA5 datasets for SiTH. While the estimated ET by MOD16 were relatively sensitive to LAI data. In addition, our results suggested that the GLOBMAP and GIMMS datasets were more suitable for long-term ET simulations than the GLASS dataset.