Extreme Learning Machine Method Research Articles

Manifold semi-supervised learning for aluminum electrolysis temperature identification based on regularized hierarchical extreme learning machine

The aim of this article is to develop a soft approach for a real-time cell temperature prediction in the aluminum electrolysis reduction. Under the limited labeled data constraint, Laplacian semi-supervised learning methods, which can fully utilize the underlying structure of the data distribution and further extract information contained in all available data, has recently received extensive attention in the field of soft sensor modeling. Since the Laplacian underlying manifold is a constant, it remains a challenging task to improve the extrapolating ability for the case that only a few labeled samples are available. This study presents a soft modeling method based on a semi-supervised deep learning structure, which was developed from the hierarchical autoencoders with extreme learning machine. Furthermore, a Laplacian–Hessian semi-supervised extreme learning machine is built to learn all the geometric distribution information. The Laplacian–Hessian semi-supervised extreme learning machine method is applied to estimate the cell temperature in an aluminum reduction process. The experimental results demonstrate the performance and robustness of the proposed algorithm are superior to those of the existing state-of-the-art methods.

SELM: Siamese extreme learning machine with application to face biometrics.

Extreme learning machine (ELM) is a powerful classification method and is very competitive among existing classification methods. It is speedy at training. Nevertheless, it cannot perform face verification tasks properly because face verification tasks require the comparison of facial images of two individuals simultaneously and decide whether the two faces identify the same person. The ELM structure was not designed to feed two input data streams simultaneously. Thus, in 2-input scenarios, ELM methods are typically applied using concatenated inputs. However, this setup consumes two times more computational resources, and it is not optimized for recognition tasks where learning a separable distance metric is critical. For these reasons, we propose and develop a Siamese extreme learning machine (SELM). SELM was designed to be fed with two data streams in parallel simultaneously. It utilizes a dual-stream Siamese condition in the extra Siamese layer to transform the data before passing it to the hidden layer. Moreover, we propose a Gender-Ethnicity-dependent triplet feature exclusively trained on various specific demographic groups. This feature enables learning and extracting useful facial features of each group. Experiments were conducted to evaluate and compare the performances of SELM, ELM, and deep convolutional neural network (DCNN). The experimental results showed that the proposed feature could perform correct classification at 97.87% accuracy and 99.45% area under the curve (AUC). They also showed that using SELM in conjunction with the proposed feature provided 98.31% accuracy and 99.72% AUC. SELM outperformed the robust performances over the well-known DCNN and ELM methods.

Short-term Wind Power Prediction Based on Soft Margin Multiple Kernel Learning Method

For short-term wind power prediction, a soft margin multiple kernel learning (MKL) method is proposed. In order to improve the predictive effect of the MKL method for wind power, a kernel slack variable is introduced into each base kernel to solve the objective function. Two kinds of soft margin MKL methods based on hinge loss function and square hinge loss function can be obtained when hinge loss functions and square hinge loss functions are selected. The improved methods demonstrate good robustness and avoid the disadvantage of the hard margin MKL method which only selects a few base kernels and discards other useful kernels when solving the objective function, thereby achieving an effective yet sparse solution for the MKL method. In order to verify the effectiveness of the proposed method, the soft margin MKL method was applied to the second wind farm of Tianfeng from Xinjiang for short-term wind power single-step prediction, and the single-step and multi-step predictions of short-term wind power was also carried out using measured data provided by alberta electric system operator (AESO). Compared with the support vector machine (SVM), extreme learning machine (ELM), kernel based extreme learning machine (KELM) methods as well as the SimpleMKL method under the same conditions, the experimental results demonstrate that the soft margin MKL method with different loss functions can efficiently achieve higher prediction accuracy and good generalization performance for short-term wind power prediction, which confirms the effectiveness of the method.

Semi-supervised learning-assisted imaging method for electrical capacitance tomography

Despite the enormous potential of the electrical capacitance tomography technology for the process industry, one of the main technological gaps and obstacles that must be overcome is the low accuracy reconstruction. To address this technical challenge, the data-dependent prior is introduced in this study, which is combined with the electrical measurement mechanism and the domain knowledge to reshape the tomographic imaging problem into a new optimization problem. The introduction of the data-dependent prior not only bridges the physical model and the advanced semi-supervised learning technique, but also improves the adaptability of the reconstruction model. A new numerical scheme that integrates the merits of the half-quadratic splitting method and the forward-backward splitting technique with a speedup strategy is proposed to solve the challenging imaging model, leading to the reduced computational burden. A new robust sparse semi-supervised extreme learning machine method that leverages both labeled and unlabeled samples and effectively handles high-dimensional image data is developed to predict the data-dependent prior, and the training is recast into a new optimization problem that is solved by a new numerical method. The novel imaging method achieves the reconstruction paradigm shift and acts as a catalyst for improving imaging quality. The evaluation of a series of reconstruction cases validates that the novel method shows satisfactory performances, outperforming both popular and classical imaging methods in terms of robustness and reconstruction accuracy.

State of health estimation for lithium-ion batteries based on temperature prediction and gated recurrent unit neural network

Accurate state of health estimation for lithium-ion batteries is crucial to ensure the safety and reliability of electric vehicles. This study presents an accurate state of health estimation method based on temperature prediction and gated recurrent unit neural network. First, the extreme learning machine method is leveraged to forecast the entire temperature variation during the constant current charging process based on randomly discontinuous short-term charging data. Next, a finite difference method is employed to calculate the raw differential temperature variation, which is then smoothed by the Kalman filter. On this basis, multi-dimensional health features are extracted from the differential temperature curves to reflect battery degradation from multiple perspectives, and six strong correlated features are selected by the Pearson correlation coefficient method. After preparing all the related health features, the gated recurrent unit neural network is exploited to predict state of health. The feasibility of the developed method is verified by comparing with other classic approaches in terms of accuracy and reliability. The experimental results demonstrate that the proposed method can effectively lead to the error of state of health within 2.28% based on only partial random and discontinuous charging data, justifying its anticipated prediction performance. • A temperature prediction model is developed to compensate incomplete information. • Extreme learning machine is exploited to predict the missing temperature curve. • Six health features are extracted from the built temperature curves. • State of health model is estimated based on gated recurrent unit neural network. • The proposed method is validated with high accuracy for state of health estimation.

An Extreme Learning Machine Based Adaptive VISMA for Stability Enhancement of Renewable Rich Power Systems

Maintaining power system stability in renewable-rich power systems can be a challenging task. Generally, the renewable-rich power systems suffer from low and no inertia due to the integration of power electronics devices in renewable-based power plants. Power system oscillatory stability can also be affected due to the low and no inertia. To overcome this problem, additional devices that can emulate inertia without adding synchronous machines can be used. These devices are referred to as virtual synchronous machines (VISMA). In this paper, the enhancement of oscillatory stability of a realistic representative power system using VISMA is proposed. A battery energy storage system (BESS) is used as the VISMA by adding an additional controller to emulate the inertia. The VISMA is designed by using Fruit Fly Optimization. Moreover, to handle the uncertainty of renewable-based power plants, the VISMA parameters are designed to be adaptive using the extreme learning machine method. Java Indonesian Power Grid has been used as the test system to investigate the efficacy of the proposed method against the conventional POD method and VISMA tuning using other methods. The simulation results show that the proposed method can enhance the oscillatory stability of the power system under various operating conditions.

Parallel cooperation search algorithm and artificial intelligence method for streamflow time series forecasting

Reliable streamflow prediction is an important productive information in the hydrology and water resources management fields. As used to forecast the nonlinear streamflow time series, the conventional artificial intelligence model may suffer from local convergence defect and fail to track the dynamic changes of the hydrological process when the model parameters and network structure are not well identified. Thus, this research develops a practical hydrological forecasting model based on parallel cooperation search algorithm (PCSA) and extreme learning machine (ELM), where the standard ELM method is chosen as the basic forecasting model, and then the PCSA method using several smaller and independent subswarms for parallel computation is used to determine satisfying input-hidden weights and hidden biases of the ELM model. The proposed model is used to forecast the nonlinear streamflow time series of several real-world hydrological stations in China. The results demonstrate that the proposed model outperforms the standard ELM model in various evaluation indicators. Thus, the key contributions of this study lie in two aspects: (1) for the first time, the parallel computing technique is developed to improve the global search ability and resources utilization efficiency of the emerging cooperation search algorithm; (2) an artificial intelligence model coupled with parallel evolutionary optimizer is proposed to improve the prediction accuracy of hydrological time series.

The extension of Moore-Penrose generalized inverse for extreme learning machine in forecasting USD/IDR exchange rate as impact of COVID-19

Since the beginning of the COVID-19 pandemic, the Indonesian economy has undergone changes which have caused the Rupiah exchange rate against the US Dollar to increase and weaken. Throughout March to September 2020, the Rupiah exchange rate depreciated by 2.75% - 4.57% with a range of Rp. 14,000 – Rp. 16,575 per US Dollar. Depreciation has a negative impact on the Indonesian economy because it makes product prices relatively cheaper for other countries. Therefore, it is necessary to do a forecast to determine the exchange rate of the Rupiah against the US Dollar in the future. Forecasting is an activity carried out using existing data to predict something in the future. In forecasting USD/IDR Exchange Rate using the Extreme Learning Machine method. This method does not require parametric assumptions and has a faster learning speed feedforward by determining the weights and biases on the network randomly. The ELM formulation leads to solving a system of linear equations in terms of unknown weights connecting the hidden layer to the output layer. The solution of this general system of linear equations is obtained using the Moore-Penrose Pseudo Inverse. The results of data analysis resulted in the optimum ELM network architecture (17-49-1), namely 17 neurons in the input layer, 49 neurons in the hidden layer, and 1 neuron in the output layer. The network is obtained from MSE which produces the smallest error value, which is 0.000338 on training data and 0.000139 on testing data. With this network, the results of forecasting USD/IDR Exchange Rate with MAPE are 0.2383951%. Forecasting results show that the Rupiah exchange rate has appreciation and is quite stable as expected by the government in strengthening the Indonesian economy during the COVID-19 pandemic.

Artificial Intelligence Based Sentiment Analysis for Health Crisis Management in Smart Cities

Smart city promotes the unification of conventional urban infrastructure and information technology (IT) to improve the quality of living and sustainable urban services in the city. To accomplish this, smart cities necessitate collaboration among the public as well as private sectors to install IT platforms to collect and examine massive quantities of data. At the same time, it is essential to design effective artificial intelligence (AI) based tools to handle healthcare crisis situations in smart cities. To offer proficient services to people during healthcare crisis time, the authorities need to look closer towards them. Sentiment analysis (SA) in social networking can provide valuable information regarding public opinion towards government actions. With this motivation, this paper presents a new AI based SA tool for healthcare crisis management (AISA-HCM) in smart cities. The AISA-HCM technique aims to determine the emotions of the people during the healthcare crisis time, such as COVID-19. The proposed AISA-HCM technique involves distinct operations such as pre-processing, feature extraction, and classification. Besides, brain storm optimization (BSO) with deep belief network (DBN), called BSO-DBN model is employed for feature extraction. Moreover, beetle antenna search with extreme learning machine (BAS-ELM) method was utilized for classifying the sentiments as to various classes. The use of BSO and BAS algorithms helps to effectively modify the parameters involved in the DBN and ELM models respectively. The performance validation of the AISA-HCM technique takes place using Twitter data and the outcomes are examined with respect to various measures. The experimental outcomes highlighted the enhanced performance of the AISA-HCM technique over the recent state of art SA approaches with the maximum precision of 0.89, recall of 0.88, F-measure of 0.89, and accuracy of 0.94.

Prediction of remaining useful life of rolling bearing based on fractal dimension and convolutional neural network

In order to predict the remaining useful life (RUL) of rolling bearings in complex environmental conditions, a bearing RUL prediction method based on fractal dimension and one-dimensional convolutional neural network (1D-CNN) is proposed. This method uses fractal dimension to characterize the degeneration process of the rolling bearing and combines the features of time domain, frequency domain, wavelet packet domain, and entropy domain. Fractal dimension provides an analytical method for characterizing the complexity of vibration signals. The features extracted from different feature domains can complement each other’s advantages, reveal the degradation state of the bearing more comprehensively and achieve better performance. Then, the percentage of the remaining life of the bearing is used as the degradation tracking index of the rolling bearing. Based on the idea of stratified sampling, the training samples and test samples are constructed, and then a 6-layer CNN is constructed to train the model. Finally, three commonly used data sets of full-life bearings are used to verify the model, namely, IEEE prognostics and health management 2012 Data Challenge, IMS dataset, and XJTU-SY dataset. The experimental results show that, on the three experimental datasets, compared to the long short-term memory network (LSTM) and the extreme learning machine (ELM) methods, the prediction effect of the RUL of the bearing based on the fractal dimension and 1D-CNN proposed in our paper is better. Its mean absolute error and root mean square error (RMSE) and mean absolute percentage error (MAPE) have been reduced, and the correlation index ( R2), adjusted_ R2, and relative accuracy (RA) have been improved, which can predict the RUL of the bearing more accurately.

Forecasting carbon futures price: a hybrid method incorporating fuzzy entropy and extreme learning machine

In this paper, we propose a novel hybrid model that extends prior work involving ensemble empirical mode decomposition (EEMD) by using fuzzy entropy and extreme learning machine (ELM) methods. We demonstrate this 3-stage model by applying it to forecast carbon futures prices which are characterized by chaos and complexity. First, we employ the EEMD method to decompose carbon futures prices into a couple of intrinsic mode functions (IMFs) and one residue. Second, the fuzzy entropy and K-means clustering methods are used to reconstruct the IMFs and the residue to obtain three reconstructed components, specifically a high frequency series, a low frequency series, and a trend series. Third, the ARMA model is implemented for the stationary high and low frequency series, while the extreme learning machine (ELM) model is utilized for the non-stationary trend series. Finally, all the component forecasts are aggregated to form final forecasts of the carbon price for each model. The empirical results show that the proposed reconstruction algorithm can bring more than 40% improvement in prediction accuracy compared to the traditional fine-to-coarse reconstruction algorithm under the same forecasting framework. The hybrid forecasting model proposed in this paper also well captures the direction of the price changes, with strong and robust forecasting ability, which is significantly better than the single forecasting models and the other hybrid forecasting models.

Data Classification Methodology for Electronic Noses Using Uniform Manifold Approximation and Projection and Extreme Learning Machine

The classification and use of robust methodologies in sensor array applications of electronic noses (ENs) remain an open problem. Among the several steps used in the developed methodologies, data preprocessing improves the classification accuracy of this type of sensor. Data preprocessing methods, such as data transformation and data reduction, enable the treatment of data with anomalies, such as outliers and features, that do not provide quality information; in addition, they reduce the dimensionality of the data, thereby facilitating the tasks of a machine learning classifier. To help solve this problem, in this study, a machine learning methodology is introduced to improve signal processing and develop methodologies for classification when an EN is used. The proposed methodology involves a normalization stage to scale the data from the sensors, using both the well-known min−max approach and the more recent mean-centered unitary group scaling (MCUGS). Next, a manifold learning algorithm for data reduction is applied using uniform manifold approximation and projection (UMAP). The dimensionality of the data at the input of the classification machine is reduced, and an extreme learning machine (ELM) is used as a machine learning classifier algorithm. To validate the EN classification methodology, three datasets of ENs were used. The first dataset was composed of 3600 measurements of 6 volatile organic compounds performed by employing 16 metal-oxide gas sensors. The second dataset was composed of 235 measurements of 3 different qualities of wine, namely, high, average, and low, as evaluated by using an EN sensor array composed of 6 different sensors. The third dataset was composed of 309 measurements of 3 different gases obtained by using an EN sensor array of 2 sensors. A 5-fold cross-validation approach was used to evaluate the proposed methodology. A test set consisting of 25% of the data was used to validate the methodology with unseen data. The results showed a fully correct average classification accuracy of 1 when the MCUGS, UMAP, and ELM methods were used. Finally, the effect of changing the number of target dimensions on the reduction of the number of data was determined based on the highest average classification accuracy.

PERBANDINGAN METODE SEASONAL ARIMA DAN EXTREME LEARNING MACHINE PADA PERAMALAN JUMLAH WISATAWAN MANCANEGARA KE BALI

Bali Island of the Gods is one of the wealth of very popular tourist destinations and has the highest number of foreign tourists in Indonesia. It is very necessary to do more in-depth learning related to the projections or forecasting of foreign tourist visits to Bali at a certain period of time. Forecasting analysis used is to compare two methods, namely the Seasonal ARIMA method (SARIMA) and Extreme Learning Machine (ELM). The SARIMA method is a statistical method commonly used in forecasting time series data that contains seasonality and has good accuracy. While the ELM method is a new learning method of artificial neural networks that has fast learning speed and good accuracy. The results obtained indicate that the Seasonal ARIMA method is a better method used to predict the number of tourists to Bali in this case, because it has a smaller forecasting MAPE value of 4.97%. While the ELM method has a forecasting MAPE value of 7.62%.

Studying the relationship between the hydraulic and geometry characteristics of labyrinth weirs based on the historical memory of reported data

One of the effective ways to increase the efficiency of weirs is to use nonlinear weirs, such as labyrinth weir, which increases the flow capacity by increasing the length of the weir at a fixed width. Given the importance of precisely estimating the flow discharge coefficient of this type of weir and its impact on supplying the safety of water structures, in the present study, the flow coefficient of labyrinth weirs was estimated using data-driven models of Extreme Learning Machine (ELM), Classification And Regression Tree (CART), Chi-square Automatic Interaction Detector (CHAID), and Multiple Linear Regression (MLR). After the modeling process, the predicted results were compared with the observed values using statistical measures and diagnostic analysis. In this study, three input combinations of hydraulic parameters, including the total upstream hydraulic head of weir (HT), weir discharge (Q), and head to weir height (HT/P) were used as input vectors. In order to evaluate the accuracy of the models, the statistical indicators of Coefficient of Efficiency (CE), RMSE, MDE, and RSD were employed. The final results showed that the ELM method created with all potential input parameters (HT, Q, and HT/P) was highly accurate in determining flow discharge coefficient. Due to having the lowest error (CE = 0.8815, RMSE = 0.0370), it was selected as the superior model.

Forecasting model of power generated by wind power plants

The power generated by wind power plants is unstable so forecasting is needed to maintain the power balance in an interconnected system. The purpose of this research is to predict the power generated at the Sidrap and Jeneponto wind power plants. The method used is an optimally pruned extreme learning machine (OPELM). The extreme learning machine (ELM) method is used as a comparison method. The mean absolute percentage error (MAPE) method is used to assess the level of forecasting accuracy. Forecasting power generation with Sidrap wind power plant data using the OPELM method is 0.8970% more accurate than the ELM which is 1.0853%. In general, the OPELM method is more accurate. Forecasting power generation with data from the Jeneponto wind power plant using the OPELM method is 2.4887% more accurate than the ELM method is 2.9984%. These results indicate that linear, sigmoid, and Gaussian activation in the OPELM method can increase accuracy. The OPELM method can be tested in forecasting the power generation at the Sidrap and Jeneponto wind power plants to maintain a power balance in the Sulselbar power grid system.

Study of power flow in electricity system using extreme learning machine

Load Flow or Power Flow Analysis in the power system in used to determine the power system parameters such as voltage, current, active power, and reactive power contained in the power grid. The method that has long been used in the calculation of load flow or power flow is the Newton-Raphson iteration method. As for its development, to complete the power flow study, it is carried out by implementing the Artificial Intelligence method, one of which is the Extreme Learning Machine method. This method is used in the simulation of the simple 39 Bus system calculation from IEEE. In this Extreme Learning Machine, the testing analysis is carried out with 2 inputs, 1 hidden layer, 5 neurons, and 2 outputs and the number of datasets is 39 to produce MAE and MAPE respectively 2.02 and 0.76% and with a very fast processing time of 0.010s

Hybrid Extreme Learning Machine dan Firefly Algorithm untuk Meramalkan Nilai Tukar Rupiah terhadap Dolar

Every country has a currency as a medium of exchange and the movement of its exchange rate can affect the economy of the country. In Indonesia, since the freely floating exchange rates system has been applied in August 1997, the value of rupiah currency in the foreign exchange market can change at any time. Considering the massive impacts of exchange rate fluctuation on the economy, then forecasting the exchange rate of rupiah against the US dollar is important to help Indonesia’s economic growth. The aims of this thesis is to predict the estimated exchange rate of rupiah against the US dollar in the future by using hybrid artificial neural network extreme learning machine (ELM) method and firefly algorithm (FA). In the training process, ELM-FA hybrid has a role to obtain the best weight and bias. The weight and bias that obtained will be used for forecasting and to know the success rate of the training process, the validation test process is required. Based on the implementation of program and simulation for some parameter values on the exchange rate data from Jan 2015 until Jan 2018, with four input and hidden nodes, and one output node, obtained the smallest MSE of the training is 0.000480513 with MSE of the testing is 0.0000854107. The relatively small MSE value indicates that ELM-FA network is able to recognize the data pattern well and able to predict the test data well.

Subset based error recovery

We propose a data denoising method using Extreme Learning Machine (ELM) structure which allows us to use Johnson-Lindenstrauß Lemma (JL) for preserving Restricted Isometry Property (RIP) in order to give theoretical guarantees for recovery. Furthermore, we show that the method is equivalent to a robust two-layer ELM that implicitly benefits from the proposed denoising algorithm. Current robust ELM methods in the literature involve well-studied L1, L2 regularization techniques as well as the usage of the robust loss functions such as Huber Loss. We extend the recent analysis on the Robust Regression literature to be effectively used in more general, non-linear settings and to be compatible with any ML algorithm such as Neural Networks (NN). These methods are useful under the scenario where the observations suffer from the effect of heavy noise. We extend the usage of ELM as a general data denoising method independent of the ML algorithm. Tests for denoising and regularized ELM methods are conducted on both synthetic and real data. Our method performs better than its competitors for most of the scenarios, and successfully eliminates most of the noise.

Design of a temperature error correction method used for meteorology and climate research

Accurate air temperature data is necessary to conducting climate research, carbon tax, and emission reduction estimation. However, the temperature error of temperature data provided by a non-aspirated multi-plate radiation shield is on the order of 1 °C due to various factors, especially solar radiation. To reduce this error, a temperature error correction method based on a method of computational fluid dynamics (CFD) and a method of extreme learning machine (ELM) for a radiation shield is proposed. The CFD method is employed to quantify the temperature errors of this type of shield. Then, the ELM method is adopted to form a universal temperature error correction equation. Finally, field experiments are performed to verify the correction accuracy. A 076B aspirated radiation shield served as a temperature reference. The root mean square error (RMSE) and the mean absolute error (MAE) between the measured temperature error results of a 41003 shield provided by the experiments and the corrected temperature error results from this correction method are 0.06 °C, and 0.056 °C, respectively. The RMSE and MAE of a 5980 shield are 0.074 °C and 0.062 °C, respectively. These results indicated that the temperature data might be expected to achieve the accuracy of 0.1 °C by using this correction method.

Constructing biomarker for early diagnosis of aMCI based on combination of multiscale fuzzy entropy and functional brain connectivity

ObjectiveTo achieve the early diagnosis of amnestic mild cognitive impairment (aMCI), this paper proposes a multi-dimensional index, which combines the advantages of the multiscale fuzzy entropy (FuzzyEn) and phase locking value (PLV) based on electroencephalography (EEG). MethodsThe complexity and synchronization of the EEG were characterized using FuzzyEn and PLV in five frequency bands, respectively. By combining the two methods, the changes in the health of brain function were comprehensively analyzed. The extreme learning machine (ELM) method was used to classify aMCI patients based on a multi-dimensional index. ResultsCompared with aMCI patients, the multiscale FuzzyEnand PLV of normal controls (NC) were higher and statistically significant (P < 0.05) in the Fp1 and Fp2channels.Moreover,significant correlation existed between the multiscale FuzzyEnor PLV and the MoCA scores in the Fp1 and Fp2 channels. The classification accuracy and running time based on ELM in the prefrontal lobe were 83.34% and 0.003 s, respectively. ConcludesThe multi-dimensional index based on prefrontal lobe could diagnosis cognitive decline of aMCI patients. SignificanceThe results showed that features integrated multiscale FuzzyEn and PLV could be used as a biomarker of cognitive decline and help realize the early diagnosis of aMCI patients.