Artificial Neural Network Hyperparameter Effectiveness Determination And Optimization Algorithm

Model Performance ◽

Activation Function ◽

Ann Model ◽

Reproductive Process ◽

Hyperparameter Selection ◽

Artificial Neural ◽

Individual Variables

Machine learning models can contain many layers and branches. Each branch and layer, contain individual variables, know as hyperparameters, that require manual tuning. For instance, the genetic algorithm designed by Unit Amin [2] was designed to mimic the reproductive process of living organisms. The genetic algorithm and the Artificial Neural Network (ANN) training processes contain inherent randomness that reduces the replicability of results. Combined with the sheer magnitude of hyperparameter permutations, confidence that model has arrived at the best solution may be low. The algorithm designed for this thesis was designed to isolate portions of a complex ANN model and generate results showing the effect each hyperparameter has on the performance of the model as a whole. The results of this thesis show that the algorithm effectively generates data correlating model performance to hyperparameter selection. This is evident in section 3.1, and 3.2, where it is shown that using the sigmoid activation function with CNN layers, regardless of the number of filters, or hyperparameters used in the subsequent LSTM layers, produces superior RMSE scores. Section 3.2 also reveals that the model does not improve in performance as the number of CNN and LSTM layers are added to the model. Finally, the results in section 3.3 show that the rmsprop optimizer generates superior results regardless of the hyperparameters used in the rest of the model.

Power Forecasting from Solar Panels Using Artificial Neural Network in UTHM Parit Raja

Journal of Advanced Industrial Technology and Application ◽

10.30880/jaita.2021.02.01.003 ◽

2021 ◽

Vol 02 (01) ◽

Author(s):

Natasha Munirah Mohd Fahmi ◽

◽

Nor Aira Zambri ◽

Norhafiz Salim ◽

Sim Sy Yi ◽

...

Keyword(s):

Neural Network ◽

Climatic Factors ◽

Activation Function ◽

Solar Panels ◽

Ann Model ◽

Power Characteristics ◽

Step Procedure ◽

Artificial Neural ◽

This paper presents a step-by-step procedure for the simulation of photovoltaic modules with numerical values, using MALTAB/Simulink software. The proposed model is developed based on the mathematical model of PV module, which based on PV solar cell employing one-diode equivalent circuit. The output current and power characteristics curves highly depend on some climatic factors such as radiation and temperature, are obtained by simulation of the selected module. The collected data are used in developing Artificial Neural Network (ANN) model. Multilayer Perceptron (MLP) and Radial Basis Function (RBF) are the techniques used to forecast the outputs of the PV. Various types of activation function will be applied such as Linear, Logistic Sigmoid, Hyperbolic Tangent Sigmoid and Gaussian. The simulation results show that the Logistic Sigmoid is the best technique which produce minimal root mean square error for the system.

The Use of an Artificial Neural Network for Predicting the Gloss of Thermally Densified Wood Veneers

BALTIC FORESTRY ◽

10.46490/bf422 ◽

2021 ◽

Vol 27 (2) ◽

Author(s):

Şükrü Özşahin ◽

Hilal Singer

Keyword(s):

Neural Network ◽

Model Performance ◽

Coefficient Of Determination ◽

Densified Wood ◽

Percentage Error ◽

Angle Of Incidence ◽

Ann Model ◽

Artificial Neural ◽

In this study, an artificial neural network (ANN) model was developed to predict the gloss of thermally densified wood veneers. A custom application created with MATLAB codes was employed for the development of the multilayer feed-forward ANN model. The wood species, temperature, pressure, measurement direction, and angle of incidence were considered as the model inputs, while the gloss was the output of the ANN model. Model performance was evaluated by using the mean absolute percentage error (MAPE), the root mean square error (RMSE), and the coefficient of determination (R²). It was observed that the ANN model yielded very satisfactory results with acceptable deviations. The MAPE, RMSE, and R2 values of the testing period of the ANN model were found as 8.556%, 1.245, and 0.9814, respectively. Consequently, this study could be useful for the wood industry to predict the gloss with less number of tiring experimental activities.

Comparative Study between Wavelet Artificial Neural Network (WANN) and Artificial Neural Network (ANN) Models for Groundwater Level Forecasting

Indian Journal of Agricultural Research ◽

10.18805/ijare.a-5079 ◽

2019 ◽

Author(s):

Ananda Kumar ◽

B Maheshwara Babu ◽

U Satish Kumar ◽

G.V Srinivasa Reddy

Keyword(s):

Neural Network ◽

Groundwater Level ◽

Model Performance ◽

Substantial Improvement ◽

Ann Model ◽

Groundwater Level Fluctuations ◽

Calibration And Validation ◽

Artificial Neural ◽

Groundwater level fluctuation modeling is a prime need for effective utilization and planning the conjunctive use in any basin.The application of Artificial Neural Network (ANN) and hybrid Wavelet ANN (WANN) models was investigated in predicting Groundwater level fluctuations. The RMSE of ANN model during calibration and validation were found to be 0.2868 and 0.3648 respectively, whereas for the WANN model the respective values were 0.1946 and 0.1695. Efficiencies during calibration and validation for ANN model were 0.8862 per cent and 0.8465 per cent respectively, whereas for WANN model were found to be much higher with the respective values of 0.9436 per cent and 0.9568 per cent indicating substantial improvement in the model performance. Hence hybrid ANN model is the promising tool to predict water table fluctuation as compared to ANN model.

Modeling and Optimization of M-cresol Isopropylation for Obtaining N-thymol: Combining a Hybrid Artificial Neural Network with a Genetic Algorithm

International Journal of Chemical Reactor Engineering ◽

10.2202/1542-6580.1605 ◽

2007 ◽

Vol 5 (1) ◽

Author(s):

Yao Kouassi Benjamin ◽

Emmanuel Assidjo Nogbou ◽

Gossan Ado ◽

Catherine Azzaro-Pantel ◽

André Davin

Keyword(s):

Neural Network ◽

Genetic Algorithm ◽

Operating Conditions ◽

Molar Ratio ◽

Ann Model ◽

Sequential Method ◽

Modeling And Optimization ◽

Artificial Neural ◽

Better Than

The application of a hybrid framework based on the combination, artificial neural network-genetic algorithm (ANN-GA), for n-thymol synthesis modeling and optimization has been developed. The effects of molar ratio propylene/cresol (X1), catalyst mass (X2) and temperature (X3) on n-thymol selectivity Y1 and m-cresol conversion Y2 were studied. A 3-8-2 ANN model was found to be very suitable for reaction modeling. The multiobjective optimization, led to optimal operating conditions (0.55 ? X1 ? 0.77; 1.773 g ? X2 ? 1.86 g; 289.74 °C ? X3 ? 291.33 °C) representing good solutions for obtaining high n-thymol selectivity and high m-cresol conversion. This optimal zone corresponded to n-thymol selectivity and m-cresol conversion ranging respectively in the interval [79.3; 79.5]% and [13.4 %; 23.7]%. These results were better than those obtained with a sequential method based on experimental design for which, optimum conditions led to n-thymol selectivity and m-cresol conversion values respectively equal to 67% and 11%. The hybrid method ANN-GA showed its ability to solve complex problems with a good fitting.

Intelligent Prediction Model of the Triaxial Compressive Strength of Rock Subjected to Freeze-Thaw Cycles Based on a Genetic Algorithm and Artificial Neural Network

Geofluids ◽

10.1155/2021/1250083 ◽

2021 ◽

Vol 2021 ◽

pp. 1-12

Author(s):

Xin Xiong ◽

Feng Gao ◽

Keping Zhou ◽

Yuxu Gao ◽

Chun Yang

Keyword(s):

Neural Network ◽

Genetic Algorithm ◽

Compressive Strength ◽

Rock Mass ◽

Ann Model ◽

Cold Regions ◽

Freeze Thaw ◽

Triaxial Compressive Strength ◽

Artificial Neural

Rock compressive strength is an important mechanical parameter for the design, excavation, and stability analysis of rock mass engineering in cold regions. Accurate and rapid prediction of rock compressive strength has great engineering value in guiding the efficient construction of rock mass engineering in a cold regions. In this study, the prediction of triaxial compressive strength (TCS) for sandstone subjected to freeze-thaw cycles was proposed using a genetic algorithm (GA) and an artificial neural network (ANN). For this purpose, a database including four model inputs, namely, the longitudinal wave velocity, porosity, confining pressure, and number of freeze-thaw cycles, and one output, the TCS of the rock, was established. The structure, initial connection weights, and biases of the ANN were optimized progressively based on GA. After obtaining the optimal GA-ANN model, the performance of the GA-ANN model was compared with that of a simple ANN model. The results revealed that the proposed hybrid GA-ANN model had a higher accuracy in predicting the testing datasets than the simple ANN model: the root mean square error (RMSE), mean absolute error (MAE), and R squared ( R 2 ) were equal to 1.083, 0.893, and 0.993, respectively, for the hybrid GA-ANN model, while the corresponding values were 2.676, 2.153, and 0.952 for the simple ANN model.

Optimization of Culture Medium for Maximal Production of Spinosad Using an Artificial Neural Network - Genetic Algorithm Modeling

Journal of Molecular Microbiology and Biotechnology ◽

10.1159/000381312 ◽

2015 ◽

Vol 25 (4) ◽

pp. 253-261 ◽

Cited By ~ 9

Author(s):

Zhou Lan ◽

Chen Zhao ◽

Weiqun Guo ◽

Xiong Guan ◽

Xiaolin Zhang

Keyword(s):

Neural Network ◽

Genetic Algorithm ◽

Ann Model ◽

Saccharopolyspora Spinosa ◽

Medium Components ◽

Key Factor ◽

Artificial Neural ◽

Artificial Neural Network Ann ◽

Rsm Model

Background: Spinosyns, products of secondary metabolic pathway of Saccharopolyspora spinosa, show high insecticidal activity, but difficulty in enhancing the spinosad yield affects wide application. The fermentation process is a key factor in this case. Methods: The response surface methodology (RMS) and artificial neural network (ANN) modeling were applied to optimize medium components for spinosad production using S. spinosa strain CGMCC4.1365. Experiments were performed using a rotatable central composite design, and the data obtained were used to construct an ANN model and an RSM model. Using a genetic algorithm (GA), the input space of the ANN model was optimized to obtain optimal values of medium component concentrations. Results: The regression coefficients (R2) for the ANN and RSM models were 0.9866 and 0.9458, respectively, indicating that the fitness of the ANN model was higher. The maximal spinosad yield (401.26 mg/l) was obtained using ANN/GA-optimized concentrations. Conclusion: The hybrid ANN/GA approach provides a viable alternative to the conventional RSM approach for the modeling and optimization of fermentation processes.

GENETIC ALGORITHM-ASSISTED ARTIFICIAL NEURAL NETWORK FOR THE ESTIMATION OF DRILLING PARAMETERS OF MAGNESIUM AZ91 IN VERTICAL MILLING MACHINE

Surface Review and Letters ◽

10.1142/s0218625x19502214 ◽

2020 ◽

Vol 27 (10) ◽

pp. 1950221

Author(s):

M. VARATHARAJULU ◽

G. JAYAPRAKASH ◽

N. BASKAR ◽

A. SARAVANAN

Keyword(s):

Neural Network ◽

Genetic Algorithm ◽

High Speed Steel ◽

Milling Machine ◽

Ann Model ◽

Drilling Parameters ◽

Vertical Milling ◽

Vertical Milling Machine ◽

Artificial Neural

The selection of appropriate drilling parameters is essential for improving productivity and part quality, therefore, this work mainly concentrates on the investigation of drilling time, burr height, burr thickness, roundness and surface roughness. The drilling experiments were carried out on Magnesium (Mg) AZ91 with High Speed Steel (HSS) tool using the Vertical Milling Machine (VMM). The parameters reckoned are spindle speed and feed rate. Artificial Neural Network (ANN) was concerned with the building of the model that will be used to forecast the responses following the consideration of Response Surface Methodology (RSM). Conventional method of modeling (RSM) yields poorer results which redirected the study with ANN. The Genetic Algorithm (GA)-based ANN has been reckoned for developing the model. With two nodes in the parameter layer and seven nodes in the response layer, six different networks were constructed using variety of nodes in the hidden layers which are 2–6–7, 2–7–7, 2–8–7, 2–6–6–7, 2–7–6–7 and 2–8–6–7. It is observed that the 2–8–7 network offers the best ANN model in predicting the various responses. The prediction results ensure the reliability of the ANN model to analyze the effect of drilling parameters over the various responses.

Stream flow forecasting by artificial neural network (ANN) model trained by real coded genetic algorithm (GA)

Journal of Groundwater Hydrology ◽

10.5917/jagh1987.48.233 ◽

2006 ◽

Vol 48 (4) ◽

pp. 233-262 ◽

Cited By ~ 3

Author(s):

Sohail A ◽

K. WATANABE ◽

S. TAKEUCHI

Keyword(s):

Neural Network ◽

Genetic Algorithm ◽

Stream Flow ◽

Ann Model ◽

Real Coded Genetic Algorithm ◽

Artificial Neural ◽

Short-term water demand predictions coupling an Artificial Neural Network model and a Genetic Algorithm

Water Science & Technology Water Supply ◽

10.2166/ws.2021.049 ◽

2021 ◽

Author(s):

Majid Gholami Shirkoohi ◽

Mouna Doghri ◽

Sophie Duchesne

Keyword(s):

Neural Network ◽

Genetic Algorithm ◽

Water Demand ◽

Urban Water ◽

Ann Model ◽

Short Term ◽

Urban Water Demand ◽

Artificial Neural ◽

Mean Square Errors

Abstract The application of artificial neural network (ANN) models for short-term (15 min) urban water demand predictions is evaluated. Optimization of the ANN model's hyperparameters with a Genetic Algorithm (GA) and use of a growing window approach for training the model are also evaluated. The results are compared to those of commonly used time series models, namely the Autoregressive Integrated Moving Average (ARIMA) model and a pattern-based model. The evaluations are based on data sets from two Canadian cities, providing 15 minute water consumption records over respectively 5 years and 23 months, with a respective mean water demand of 14,560 and 887 m3/d. The GA optimized ANN model performed better than the other models, with Nash-Sutcliffe Efficiencies of 0.91 and 0.83, and Relative Root Mean Square Errors of 6 and 16% for City 1 and City 2, respectively. The results of this study indicate that the optimization of the hyperparameters of an ANN model can lead to better 15 min urban water demand predictions, which are useful for many real time control applications, such as dynamic pressure control.