Maximum Negative Error Research Articles

High correlated variables creator machine: Prediction of the compressive strength of concrete

In this paper, we introduce a novel hybrid model for predicting the compressive strength of concrete using Ultrasonic Pulse Velocity (UPV) and Rebound Number (RN). First, we collect 516 datasets from 8 studies of UPV and Rebound Hammer (RH) tests. Then, we propose the High Correlated Variables Creator Machine (HCVCM) to create the new variables that have a better correlation with the output in order to improve the prediction models. Three single models, including a Step-By-Step Regression (SBSR), Gene Expression Programming (GEP), an Adaptive Neuro-Fuzzy Inference System (ANFIS) as well as three hybrid models, i.e. HCVCM-SBSR, HCVCM-GEP, and HCVCM-ANFIS are employed to predict the compressive strength of concrete. The statistical parameters and error terms such as the coefficient of determination, the Root Mean Square Error (RMSE), Normalized Mean Square Error (NMSE), fractional bias, the maximum positive and negative errors, and the Mean Absolute Percentage Error (MAPE) are computed to evaluate the models. The results show that HCVCM-ANFIS can predict the compressive strength of concrete better than all other models. HCVCM improves the accuracy of ANFIS by 5% in the coefficient of determination, 10% in the RMSE, 3% in the NMSE, 20% in MAPE, and 7% in the maximum negative error.

Modelling the power output from a steam power plant in Nigeria

Adequate supply of electricity at a competitive price is pivotal to sustainable development. More often than not, the generation of electricity which drives modern growth and development is currently powered by limited fossil fuels in many nations. Electricity generation and megawatt demand are also usually fluctuating due to several pertinent factors. In a bid to articulate the impact of inherent variations in process parameters on the performance of steam power plant at different loads, this paper presents an investigation into the efficacy of two validation strategies in predicting the net power output from the plant using GMDH Shell software. Using the combinatorial algorithm, the k-fold cross-validation strategy and the training/testing validation technique were applied to empirical data of a power plant in Nigeria. The performance of the models returned from the two validation strategies was evaluated using maximum negative error, maximum positive error, mean absolute percentage error (MAPE), root mean square percentage error (RMSPE), residual sum, the standard deviation of residuals, coefficient of determination (R2) and correlation. For the number of folds and the training/testing split percentage considered in this study, results show that both models obtained were quite competitive, with the k-fold model having a slight edge over the other model. It is expected that the outcome of the study will be handy in researches for providing knowledge base information on choosing and setting optimum operating conditions at various load demand.

Critical Analysis of “Flip-Flop” Phenomenon in Two‐Compartment Pharmacokinetic Model

Computer simulations were used to examine the effect of first‐order absorption on the disposition of one‐ and two‐compartment model drugs. Two‐compartment systems that attain a clinically acceptable β‐phase after rapid intravenous injection were perturbed by introduction of drug via first‐order absorption. The validity of perceiving such a system as a potential “flip‐flop” model was tested by comparing the negative slopes of log‐linear plasma‐time profiles to known values for ka and β for various values of ka, k12, k21, and k21. Although most log‐linear plots showed excellent correlation coefficients (r2 > 0.996), their negative slopes (S) did not represent either k1 or β under various combinations. A similar consideration of the one‐compartment model enabled a comparison to be made between the two systems. Maximum negative errors were observed for both one‐ and two‐compartment drugs as ka → k2 or β, respectively. The value for S provided a good estimate of the absorption rate constant, ka, when k2 ≥ 2ka (one compartment) or β ≥ 2ka. The elimination rate constant (k2 or β) could be obtained from S for all one‐compartment and some two‐compartment drugs when the value of ka was approximately twice that of k2 or β. Large positive errors also were observed with certain two‐compartment drugs where the ratio of the four rate constants apparently linearized a nonlinear plasma profile. Conditions wherein S may be expected to approach β wherein S approaches ka are clearly defined.