Root Mean Square Error Percentage Research Articles

Artificial neural networks based prediction of hourly horizontal solar radiation data: case study

The aim of the present study is to predict global solar radiation (GSR) received on the horizontal surface using artificial neural network (ANN). The measured data of the year (2013) was provided by the Applied Research Unit of Ghardaia - Algeria. The best results were obtained with a 7/24/1 ANN model trained with the quasi-Newton back propagation (BFGS) algorithm. The prediction accuracy for the internal and the external validation set was estimated by the Q2LOO and Q2ext which are equal to 0.9984, 0.9977 for ANN, with percent root mean square error (PRMSE) of 4.71% and the mean bias error (MBE) 0.021% for the internal validation and 5.60%, 0.42% for the external validation, respectively. These results show that the optimised model is robust and have a good predictive power explained by a good agreement between the measurement and prediction values of the solar radiation.

State space versus SARIMA modeling of the Nigeria’s crude oil export

This paper analyzes the Nigeria’s crude oil export series using monthly data from January 1999 to December 2014. We employed the state space local level model with stochastic and deterministic seasonal to model the dynamic features in the Nigeria crude oil export. Our results clearly indicate that the local level model with deterministic seasonal is the most parsimonious model between the two state space models considered in this study. Also, a parsimonious SARIMA model is also fitted to the data. We compare the forecasting performance of the two parsimonious models and evaluate their forecasts using ex-post indicators such as mean absolute percentage error (MAPE), root mean square percentage error (RMSPE) and the Theil’s U statistic. The forecast analysis and evaluation results indicate that the state space local level model with deterministic seasonal outperforms the Box-Jenkins model in shorter and medium – range forecasting horizons. Howbeit, the forecast of the SARIMA model improves in the longer horizon. The Theil’s U statistic also indicates that the state space local level model with deterministic seasonal and SARIMA model outperform the naive model at most of the forecasting horizons. In conclusion, we recommend that the state space model with deterministic seasonal component should be used in shorter and medium range forecasting horizons of the Nigeria’s monthly crude oil export. Howbeit, for longer forecasting horizon, ten months and above, the seasonal ARIMA model should be considered.

Allometric equations for estimating tree aboveground biomass in evergreen broadleaf forests of Viet Nam

For mitigating climate change through carbon sequestration and for reporting, Viet Nam needs to develop biomass equations at a national scale. These equations need to be accurate and provide quantifiable uncertainty. Using data from 968 trees across five ecoregions of Viet Nam, we developed a set of models to estimate tree aboveground biomass (AGB) in evergreen broadleaf forests (EBLF) at the national level. Diameter at breast height (DBH), tree height (H), wood density (WD), and combination of these three tree characteristics were used as covariates of the biomass models. Effect of ecoregion, wood density, plant family on AGB were examined. Best models were selected based on AIC, Adjusted R2, and visual interpretation of model diagnostics. Cross-validation statistics of percent bias, root mean square percentage error (RMSPE), and mean absolute percent error (MAPE) were computed by randomly splitting data 200 times into model development (80%) and validation (20%) datasets and averaging over the 200 realizations. Effects models were used, the best results were obtained by using a combined variable (DBH2HWD (kg)=(DBH (cm)/100)2×H (m)×WD (g/cm3)×1000) model AGB=a×(DBH2HWD)b. Including a categorical WD variable as a random effect reduced AIC, percent bias, RMSPE, MAPE of models AGB=a×DBHb and AGB=a×(DBH2H)b; ecoregion as a random effect reduced the AIC of models AGB=DBHb×WD, AGB=a×(DBH2H)b, and AGB=a×(DBH2HWD)b. For models that did not include WD variable, including plant family as a random effect reduced AIC, RMSE, and MAPE; recommendations are provided for models with specific parameters for main families and without WD if this variable is not available. The overall best model for estimating AGB was the equation form AGB=a×(DBH2HWD)b with ecoregion as a random effect.

PERAMALAN JUMLAH KUNJUNGAN WISATAWAN AUSTRALIA YANG BERKUNJUNG KE BALI MENGGUNAKAN MODEL TIME VARYING PARAMETER (TVP)

The purpose of this research is to forecast the number of Australian tourists arrival to Bali using Time Varying Parameter (TVP) model based on inflation of Indonesia and exchange rate AUD to IDR from January 2010 – December 2015 as explanatory variables. TVP model is specified in a state space model and estimated by Kalman filter algorithm. The result shows that the TVP model can be used to forecast the number of Australian tourists arrival to Bali because it satisfied the assumption that the residuals are distributed normally and the residuals in the measurement and transition equations are not correlated. The estimated TVP model is . This model has a value of mean absolute percentage error (MAPE) is equal to dan root mean square percentage error (RMSPE) is equal to . The number of Australian tourists arrival to Bali for the next five periods is predicted: ; ; ; ; and (January - May 2016).

Aboveground biomass equations for evergreen broadleaf forests in South Central Coastal ecoregion of Viet Nam: Selection of eco-regional or pantropical models

As part of Viet Nam’s effort to participate in REDD+ (reducing emissions from deforestation and forest degradation), selected biomass equations were evaluated for their predictive abilities using data collected from destructively sampled 110 trees from 41 species of the evergreen broadleaf forests of the South Central Coastal region of Viet Nam. Different power models that used diameter at breast height (DBH), tree height (H), wood density (WD), and crown area (CA) as covariates to predict aboveground biomass (AGB) were evaluated. Best models were selected based on the coefficient of determination (R2), the Akaike information criterion (AIC), and root mean square percent error (RMSE). AGB was strongly related to four covariates - DBH, H, WD, and CA. While seldom mentioned in the existing literature, CA improved the accuracy of the AGB estimation. Accuracy of the selected models was validated using the random validation dataset and the model with four explanatory variables (AGB=a×(DBH2HWD)b×CAc) had the lowest mean absolute percent error of 16.9%. Using local data, a simple power model based on DBH only (AGB=a×DBHb) produced higher accuracy than the generic pantropical models that used up to three variables.

An IPCC-Compliant Technique for Forest Carbon Stock Assessment Using Airborne LiDAR-Derived Tree Metrics and Competition Index

This study developed an IPCC (Intergovernmental Panel on Climate Change) compliant method for the estimation of above-ground carbon (AGC) in forest stands using remote sensing technology. A multi-level morphological active contour (MMAC) algorithm was employed to obtain tree-level metrics (tree height (LH), crown radius (LCR), competition index (LCI), and stem diameter (LDBH)) from an airborne LiDAR-derived canopy height model. Seven biomass-based AGC models and 13 volume-based AGC models were developed using a training dataset and validated using a separate validation dataset. Four accuracy measures, mean absolute error (MAE), root-mean-square error (RMSE), percentage RMSE (PRMSE), and root-mean-square percentage error (RMSPE) were calculated for each of the 20 models. These measures were transformed into a new index, accuracy improvement percentage (AIP), for post hoc testing of model performance in estimating forest stand AGC stock. Results showed that the tree-level AGC models explained 84% to 91% of the variance in tree-level AGC within the training dataset. Prediction errors (RMSEs) for these models ranged between 15 ton/ha and 210 ton/ha in mature forest stands, which is equal to an error percentage in the range 6% to 86%. At the stand-level, several models achieved accurate and reliable predictions of AGC stock. Some models achieved 90% to 95% accuracy, which was equal to or superior to the R-squared of the tree-level AGC models. The first recommended model was a biomass-based model using the metrics LDBH, LH, and LCI and the others were volume-based models using LH, LCI, and LCR and LDBH and LH. One metric, LCI, played a critical role in upgrading model performance when banded together with LH and LCR or LDBH and LCR. We conclude by proposing an IPCC-compatible method that is suitable for calculating tree-level AGC and predicting AGC stock of forest stands from airborne LiDAR data.

Fuzzy Time Series Theory Application for the China Containerized Freight Index

China has evolved into one of the world’s largest trading nations. China has adequate supply for imports and exports, and therefore, major shipping companies from various countries around the world all joined this market to perform freight transport. Currently, the main method of transporting goods is via shipping. China’s containerized freight index (CCFI) is mainly used as a reference to evaluate the current freight tariffs standard. This study uses fuzzy time series to predict the CCFI. The results of our analysis found the following: 1. CCFI yield series has a volatility-clustering characteristic (the mean of the current yield is negative); 2. the R.M.S.P.E. (root mean square percentage error) value is 0.078%, indicating that the goodness-of-fit of the model is quite good; 3. future CCFI will be maintained at a low point of around 893.557, which is an optimistic long-term indication for freight; 4. currently, the supply of ships outweighs the demand, causing a long-term low CCFI. These four conclusions are hoped to serve as references for relevant policymakers in the future.

Analysis of DWT signal denoising on various biomedical signals by neural network

A detailed analysis of Discrete Wavelet Transform (DWT) denoising and identification on various wavelet families and biomedical signals (ECG, EEG and EMG) is presented in this paper. The main intention of this work is to explore the wavelet function which is optimal for denoising the signals. Nevertheless, wavelet transforms offer better results for denoising biomedical signals, but identification is a crucial process. This paper proposes an artificial neural network in which the wavelet types are used to denoise the signals optimally by using a learning back propagation algorithm. Also the, performances of the various wavelet types are tabulated and compared with the existing techniques, in terms of the evaluation parameters signal to noise ratio, percent root mean square difference, mean-square error and compression ratio. The simulation results expose the efficiency of the proposed method for the denoising of biomedical signals.

Evaluation of the Geostationary Ocean Color Imager (GOCI) to monitor the dynamic characteristics of suspension sediment in Taihu Lake

To accurately observe the dynamic concentration of total suspended matter (CTSM) by geostationary satellite, Geostationary Ocean Color Imager (GOCI), a new, regionally empirical combination algorithm, was developed based on the in situ measurement of water properties. Also, the ability of GOCI (with a 1 h temporal resolution) to capture dynamic change in CTSM was evaluated using in situ buoy-measured data. The validation results indicate that the combination algorithm (combination of two- and three-band ratio algorithms) can achieve acceptable retrieval results for CTSM, with 34.52% root mean square error percentage (RMSP) for validation data and 45.01% RMSP for GOCI match-up points. The depth of reflectance valley at 680 nm (Rrs-depth,680) is considered as an index for deciding whether to utilize a two- (Rrs-depth,680 ≤ 0.005) or three-band (Rrs-depth,680 > 0.005) algorithm in the combination algorithm. This combination algorithm overcomes some shortcomings of two- and three-band ratio algorithms of CTSM in turbid productive water. The results show that sediment resuspension will significantly change the optical properties of water and decrease the universality of the CTSM retrieval algorithm in the presence of a large area of algae. The influence of sediment resuspension on the CTSM retrieval algorithm is significant in the central lake but less so in the other four lakes studied due to lake morphology and algal biomass.

The Effect of Swing Arms Posture Motion on Walking Straightness of Humanoid Robot

The study of humanoid robot locomotion has been focusing on the development of walking pattern generation. However, the robot has the possibility of slipping on the ground due to the simultaneous occurrence of translational and rotational motions as a robot is walking. Technically, both issues disrupt the walking straightness of the robot. This study investigates the influence of the robot swing arms posture with a modified walking pattern generation to minimize the problem. Simulations were done by preparing 10 different elbow joint angles with swing arms and each one of them were tested with the same value of Maximum Lifting Height, H. The analysis was done by collecting the data from the Global Positional Sensor (GPS) from the robot and calculating the Root Mean Square Error (RMSE) to determine which variable has the lowest percentage of RMSE. The angular velocity graph was used to analyze the deviation encountered by the robot while walking by determining the average of angular velocity. The results showed that the arm posture of the robot does influence its walking straightness in a small amount.

Optimal estimator for uncertainty-based measurement quality control

This paper presents an optimal estimator for uncertainty-based measurement quality control. The optimal estimator is developed based on an acceptance probability approach under a risk balance criterion. It yields a balance between the false acceptance and false rejection when the measurement quality index is equal to unity. This paper also presents a minimum mean absolute percentage error (MAPE) estimator and a minimum mean squared percentage error estimator based on the frequentist decision-theoretic approach. The mathematical formulations for computing the MAPE, relative bias error (RBE), relative precision error (RPE), and root mean squared percentage error (RMSPE) of the three presented estimators and five existing estimators are derived. The performance of these estimators are compared and evaluated in terms of the false acceptance/rejection probability, MAPE, RBE, RPE, and RMSPE. Among the eight estimators considered, the presented optimal estimator is the most meaningful and the best estimator from the measurement quality control perspective.

Gasoline Price Forecasting: An Application of LSSVM with Improved ABC

Optimizing the hyper-parameters of Least Squares Support Vector Machines (LSSVM) is crucial as it will directly influence the predictive power of the algorithm. To tackle such issue, this study proposes an improved Artificial Bee Colony (IABC) algorithm which is based on conventional mutation. The IABC serves as an optimizer for LSSVM. Realized in gasoline price forecasting, the performance is guided based on Mean Absolute Percentage Error (MAPE) and Root Mean Square Percentage Error (RMSPE). The conducted simulation results show that, the proposed IABC- LSSVM outperforms the results produced by ABC-LSSVM and also the Back Propagation Neural Network.

Evaluation of heterogeneity statistics as reasonable proxies of the error of precipitation quantile estimation in the Minneapolis-St. Paul region

Summary Estimating precipitation frequency is important in engineering, agriculture, land use planning, and many other disciplines. The index flood method alleviates small sample size issues due to short record length by calculating normalized quantile estimates for averaged data from a “region” of gauges. For a perfectly homogeneous region this adds no error; heterogeneity statistics seek to quantify a real-world region’s deviation from this assumption. Hosking and Wallis (1997) introduced a Monte Carlo heterogeneity statistic called here H 1 and used a simulation study to assess its utility while rejecting two similar statistics called here H 2 and H 3 . A nearly linear relationship was found between H 1 and the percentage root mean square error (RMSE) increase due to heterogeneity, establishing H 1 as a “reasonable proxy” of quantile error. The H 1 -percent RMSE added relationship found in the simulation experiment was used to find equivalent RMSEs for heterogeneity thresholds against which all three H statistics were tested. In this study the “reasonable proxy” relationship is evaluated across a highly skewed daily precipitation dataset in Minnesota for H 1 , H 2 and H 3 . Simulated regions used in quantile error estimation are generated using at-site L-moment ratios scaled toward the regional mean with a shrinkage multiplier. A linear relationship is found between Monte Carlo estimates of quantile RMSE and both H 1 and H 2 across all possible regionalizations of twelve gauges. H 2 ’s relationship is less linear than H 1 ’s as quantified by Pearson’s r. A synthetic study is also undertaken using the same sample sizes, regional L-moment averages, and between-site variations as the Hosking and Wallis (1997) simulation. The H 2 -percent RMSE added relationship is found to be nearly as linear as for H 1 , complementing the enumeration study’s findings. Because H 2 ’s linear relationship with percent RMSE added has approximately one-fourth the slope of the H 1 -RMSE relationship, heterogeneity thresholds calculated with reference to H 1 should not be applied to H 2 . H 2 thresholds can be derived from the H 2 -percent RMSE added relationship in analogous fashion to the method used in Hosking and Wallis (1997) for H 1 . The resulting thresholds are one-fourth the magnitude of the H 1 thresholds.

Modeling Conditional Covariances With Economic Information Instruments

We propose a new model for conditional covariances based on predetermined idiosyncratic shocks as well as macroeconomic and own information instruments. The specification ensures positive definiteness by construction, is unique within the class of linear functions for our covariance decomposition, and yields a simple yet rich model of covariances. We introduce a property, invariance to variate order, that assures estimation is not impacted by a simple reordering of the variates in the system. Simulation results using realized covariances show smaller mean absolute errors (MAE) and root mean square errors (RMSE) for every element of the covariance matrix relative to a comparably specified BEKK model with own information instruments. We also find a smaller mean absolute percentage error (MAPE) and root mean square percentage error (RMSPE) for the entire covariance matrix. Supplementary materials for practitioners as well as all Matlab code used in the article are available online.

Calibration and Evaluation of Link Congestion Functions: Applying Intrinsic Sensitivity of Link Speed as a Practical Consideration to Heterogeneous Facility Types within Urban Network

This paper explores the use of archived data to calibrate volume delay functions (VDFs) and updates their input parameters (capacity and free-flow speed) for planning applications. The sensitivity analysis of speed to change in congestion level is performed to capture functional characteristics of VDFs in modeling specific facility types. Different sensitivity characteristics shown by the VDFs indicate that each function is suitable to a particular facility type. The results of sensitivity analysis are confirmed by the root mean square percent error (RMSPE) values calculated using the Orlando Urban Area Transportation Study (OUATS) model results and observed data. The modified Davidson’s function exhibits remarkable performance in nearly all facility types. The strength of the modified Davidson’s function across a broad range of facilities can be attributed to the flexibility of its tuning parameter, μ. Fitted Bureau of Public Road (BPR) and conical delay functions show lower RMSPE for uninterrupted flow facilities (freeways/expressways, managed lanes) and higher values for toll roads (which might have partial interruptions due to toll booths) and signalized arterials. Akcelik function underperforms on freeways/expressways and managed lanes but shows some improvements for toll roads and superior results for the signalized arterials. This was a desired strength of Akcelik function when modeling link travel speed on facilities where stopped delays were encountered.

Forest Attributes Estimation Using Aerial Laser Scanner and TM Data

Aim of study: The aim of this study was performance of four non-parametric algorithms including the k-NN, SVR, RF and ANN to estimate forest volume and basal area attributes using combination of Aerial Laser Scanner and Landsat-TM data.Area of study: Data in small part of a mixed managed forest in the Waldkirch region, Germany.Material and methods: The volume/ha and basal area/ha in the 411 circular plots were estimated based on DBH and height of trees using volume functions of study area. The low density ALS raw data as first and last pulses were prepared and automatically classified into vegetation and ground returns to generate two fine resolution digital terrain and surface models after noise removing. Plot-based height and density metrics were extracted from ALS data and used both separated and combined with orthorectified and processed TM bands. The algorithms implemented with different options including k-NN with different distance measures, SVR with the best regularized parameters for four kernel types, RF with regularized decision tree parameters and ANN with different types of networks. The algorithm performances were validated using computing absolute and percentage RMSe and bias on unused test samples.Main results: Results showed that among four methods, SVR using the RBF kernel could better estimate volume/ha with lower RMSe and bias (156.02 m3 ha–1 and 0.48, respectively) compared to others. In basal area/ha, k-NN could generate results with similar RMSe (11.79 m3 ha–1) but unbiased (0.03) compared to SVR with RMSe of 11.55 m3 ha–1 but slightly biased (–1.04).Research highlights: Results exposed that combining Lidar with TM data could improve estimations compared to using only Lidar or TM data.Key words: forest attributes estimation; ALS; TM; non-parametric algorithms.

Modeling and Forecasting Inbound Tourism Demand for Long-Haul Markets of Beijing

This paper aims to identify the most influencing factors of Beijing's inbound tourism demand using the autoregressive distributed lag model (ADLM) and then generates forecasts of international tourist arrivals from the United States, the United Kingdom, and Canada for the period of 2010Q3–2015Q4. The general-to-specific modeling approach was adopted to achieve final models while the exponential smoothing method was used to produce forecasts for independent variables. Results show that factors such as “word of mouth” effect, income level of the origin source markets, the costs of tourism in Beijing, and the cost of tourism in the competing destinations are crucial determinants of the tourism flows from three long-haul international markets. A group of error measures, such as the mean absolute percentage error (MAPE), root mean square percentage error (RMSPE), mean absolute error (MAE), root mean square error (RMSE), and Theil's U statistic, were used to evaluate the forecasting accuracy. The results suggest that all three models have good forecasting abilities with the MAPEs ranging from 5.73% to 14.89%. Implications are discussed and recommendations as well as future research directions are provided.

Improving Forecasting Accuracy by Combining Statistical and Judgmental Forecasts in Tourism

This study aims to examine the forecasting accuracy of a combined method by using quarterly tourism arrival data in Hong Kong. The voluntary integration of statistical forecasts and experts’ judgmental revisions are achieved through a Delphi procedure in a Web-based Tourism Demand Forecasting System (TDFS). The forecasting performance is evaluated using the absolute percentage error (APE), mean absolute percentage error (MAPE) and root mean square percentage error (RMSPE). This study also compares the forecast performance of the combined method to a number of alternative forecasting models, i.e. the Naïve models, exponential smoothing and Box-Jenkins time-series models. The empirical results show that the combined forecasts consistently outperform the baseline forecasts obtained from vector autoregression (VAR) models over the forecasting period of 2008Q1–2011Q4, suggesting the value of adopting such an integration procedure. The findings indicate that some gains are obtained from integrating experts’ judgments into statistical forecasts, particularly for the short term. In addition, combined forecasting does not always lead to satisfactory forecasting performance, particularly when there is a lack of important contextual information.

Growth modelling including size heterogeneity: Application to the intensive culture of white shrimp (P. vannamei) in freshwater

This article models the growth of a population in aquaculture, including the phenomenon of size heterogeneity. Experimental observations of six initial densities (90, 130, 180, 230, 280 and 330shrimpsm−2) in an intensive culture of Penaeus vannamei in freshwater were used to fit a growth model. For this, three mathematical functions were analysed (Gompertz, von Bertalanffy and Pütter), which were modified in order to include the effect of initial density. Two models were constructed, one assuming the hypothesis of homogeneous individual growth and the other including size-heterogeneity throughout the culture cycle. For the second case, a stabilisation time was evaluated which defines the future heterogeneity of the cohort. In each initial density, the stabilisation phase was reached at approximately 2g. However, the time taken for this phase to start increased with increasing initial culture density. The modified von Bertalannfy function was the most effective of the three equations in predicting growth. The weight predictions, assuming homogeneity and variability of sizes, presented low Percentage Root Mean Square Errors (PRMSE). However, the inclusion of size variability in the fit produced better statistical results than when they were not included (PRMSE decreased from 16% to 12%). This confirms the importance of taking into account size heterogeneity at an intensive level. Particularly at the technical level when the aquaculture producer is required to make stocking and harvest plans.

Growth-Competition-Based Stem Diameter and Volume Modeling for Tree-Level Forest Inventory Using Airborne LiDAR Data

An individual tree within a forest stand will have its height and diameter growth restricted by the influence of neighboring trees. This is because trees in close proximity compete for resources and space to enable growth. In this paper, the position of trees, tree height (LH), tree crown radius (LCR), and growth competition index (LCI) were extracted from a light-detection-and-ranging (LiDAR)-based rasterized canopy height model using the multilevel morphological active-contour algorithm. The diameter and volume of individual trees are tested and validated to be an exponential function of those LiDAR-derived tree parameters. The best LiDAR-based diameter estimation model and volume estimation model were tested as significant with an <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">R</i> <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> value of 0.84 and 0.9 and evaluated with an estimation bias of 8.7 cm and 0.91 m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> , respectively. Results also showed that LH and LCR are positively related to the LiDAR-derived diameter at breast height (DBH) and the LiDAR-derived volume of individual trees in a forest stand, whereas LCI is negatively related. The proposed algorithm of individual tree volume estimation was further applied to predict the volume of three sample plots in mountainous forest stands. It was found that the LVM could be used to predict an acceptable volume estimate of old-aged forest stands. The estimation bias, i.e., percentage RMSE (RMSE%), is averaged at around 4% using the LiDAR metrics lnLH, LCI, and LCR, whereas the RMSE% increases to 50% if only lnLH is applied. Results suggest that LCI is an important regulation factor in the estimation of forest volume stocks using LiDAR remote sensing.