Extraction of Phenolic Compounds with Antioxidant Activity from Strawberries: Modelling with Artificial Neural Networks (ANNs)

This research study focuses on the evaluation of the total phenolic compounds (TPC) and antioxidant activity (AOA) of strawberries according to different experimental extraction conditions by applying the Artificial Neural Networks (ANNs) technique. The experimental data were applied to train ANNs using feed- and cascade-forward backpropagation models with Levenberg-Marquardt (LM) and Bayesian Regulation (BR) algorithms. Three independent variables (solvent concentration, volume/mass ratio and extraction time) were used as ANN inputs, whereas the three variables of total phenolic compounds, DPPH and ABTS antioxidant activities were considered as ANN outputs. The results demonstrate that the best cascade- and feed-forward backpropagation topologies of ANNs for the prediction of total phenolic compounds and DPPH and ABTS antioxidant activity factors were the 3-9-1, 3-4-4-1 and 3-13-10-1 structures, with the training algorithms of trainlm, trainbr, trainlm and threshold functions of tansig-purelin, tansig-tansig-tansig and purelin-tansig-tansig, respectively. The best R2 values for the predication of total phenolic compounds and DPPH and ABTS antioxidant activity factors were 0.9806 (MSE = 0.0047), 0.9651 (MSE = 0.0035) and 0.9756 (MSE = 0.00286), respectively. According to the comparison of ANNs, the results showed that the cascade-forward backpropagation network showed better performance than the feed-forward backpropagation network for predicting the TPC, and the FFBP network, in predicting the DPPH and ABTS antioxidant activity factors, had more precision than the cascade-forward backpropagation network. The ANN technique is a potential method for estimating targeted total phenolic compounds and the antioxidant activity of strawberries.

A Hybrid Approach Combining Fuzzy c-Means-Based Genetic Algorithm and Machine Learning for Predicting Job Cycle Times for Semiconductor Manufacturing

Job cycle time is the cycle time of a job or the time required to complete a job. Prediction of job cycle time is a critical task for a semiconductor fabrication factory. A predictive model must forecast job cycle time to pursue sustainable development, meet customer requirements, and promote downstream operations. To effectively predict job cycle time in semiconductor fabrication factories, we propose an effective hybrid approach combining the fuzzy c-means (FCM)-based genetic algorithm (GA) and a backpropagation network (BPN) to predict job cycle time. All job records are divided into two datasets: the first dataset is for clustering and training, and the other is for testing. An FCM-based GA classification method is developed to pre-classify the first dataset of job records into several clusters. The classification results are then fed into a BPN predictor. The BPN predictor can predict the cycle time and compare it with the second dataset. Finally, we present a case study using the actual dataset obtained from a semiconductor fabrication factory to demonstrate the effectiveness and efficiency of the proposed approach.

Radar Working State Recognition Based on Improved HPSO-BP

In this paper, a recognition model based on the improved hybrid particle swarm optimisation (HPSO) optimised backpropagation network (BP) is proposed to improve the efficiency of radar working state recognition. First, the model improves the HPSO algorithm through the nonlinear decreasing inertia weight by adding the deceleration factor and asynchronous learning factor. Then, the BP neural network’s initial weights and thresholds are optimised to overcome the shortcomings of slow convergence rate and falling into local optima. In the simulation experiment, improved HPSO-BP recognition models were established based on the datasets for three radar types, and these models were subsequently compared to other recognition models. The results reveal that the improved HPSO-BP recognition model has better prediction accuracy and convergence rate. The recognition accuracy of different radar types exceeded 97%, which demonstrates the feasibility and generalisation of the model applied to radar working state recognition.

Prediksi Tinggi Permukaan Air Waduk Menggunakan Artificial Neural Network Berbasis Sliding Window

The water level in the reservoir is an important factor in the operation of a hydroelectric turbine to control water overflow so that there is no excessive degradation. This water control has an influence on the performance and production of hydroelectric energy. The daily reservoir water level (tpaw) recording of PLTA Riam Kanan is carried out through a daily direct measurement and observation process on the reservoir measuring board which is recapitulated every month in excel form. This time series historical data continues to grow every day to become a data warehouse that is still useless if only stored. Extracting knowledge from the data warehouse can be done using one of the artificial neural network data mining techniques, namely backpropagation to predict the next day's tpaw. Historical data for the tpaw time series is presented with a sliding window concept approach based on the window sizes used, namely 7, 14, 21 and 28. The window size represents the number of days as an input layer variable in the backpropagation network architecture to predict the next day's tpaw. Some backpropagation network testing is carried out using a combination of the number of window sizes against the comparison of the amount of training data and test data on the network. The prediction results obtained with the smallest mean squared error (mse) in network testing is 0.000577 as a high accuracy value of the prediction results. The network architecture with the smallest mse using 28 input layers, 10 hidden layers and 1 output layer can be a knowledge that can help the hydropower plant as an alternative in making turbine operation decisions based on the predicted results of reservoir water level.

Optimising Self-Similarity Network Traffic for Better Performance

Given the ubiquity of the burstiness present across many networking facilities and services, predicting and managing self-similar traffic has become a key issue owing to new complexities associated with self-similarity which makes difficult the achievement of high network performance and quality of service (QoS). In this study ANN model was used to model and simulate FCE Okene computer network traffic. The ANN is a 2-39-1 Feed Forward Backpropagation network implemented to predict the bursty nature of network traffic. Wireshark tools that measure and capture packets of network traffic was deployed. Moreover, variance-time method is a log-log scale plot, representing variance versus a non-overlapping block of size m aggregate variance level engaged to established conformity of the ANN approach to self-similarity characteristic of the network traffic. The predicted series were then compared with the corresponding real traffic series. Suitable performance measurements used were the Means Square Error (MSE) and the Regression Coefficient. Our results showed that burstiness is present in the network across many time scales. The study also established the characteristic property of a long-range dependence (LRD). The work recommended that network traffic observation should be longer thereby enabling larger volume of traffic to be capture for better accuracy of traffic modelling and prediction.

Performance Assessment of a Triangular Integrated Collector Using Neural Networks

A numerical study is achieved on a new shape of temperature saver solar collector using an artificial neural network. The storage collector is a triangle face and a right triangle pyramid for the volumetric shape. It is obtained by cutting a cube from one upper corner at 45°, down to the opposite hypotenuse of the base of the cube. The numerical study was carried out using the computational fluid dynamics code (ANSYS-Fluent) software with natural convection phenomenon in the pyramid enclosure. Elman backpropagation network is used for his ability to find the nearest solution with the smallest error rate. The network consists of three layers, each of different corresponding weights. The results show that the temperature and velocity distributions throughout the operating period were obtained. The influence of introducing an internal partition inside the triangular storage collector was investigated. Also the optimum geometry and location for this partition were obtained. The enhancement was best at y = 0.25 m, whereas the height of triangular collector was 0.5 m. The hourly system performance was evaluated for all test conditions. The performance of the NN to train a model for this work was 0.000207, while the error of the calculation was 1×10-2 as average.

Performance of Levenberg-Marquardt Algorithm in Backpropagation Network Based on the Number of Neurons in Hidden Layers and Learning Rate

One of the supervised learning paradigms in artificial neural networks (ANN) that are in great developed is the backpropagation model. Backpropagation is a perceptron learning algorithm with many layers to change weights connected to neurons in hidden layers. The performance of the algorithm is influenced by several network parameters including the number of neurons in the input layer, the maximum epoch used, learning rate (lr) value, the hidden layer configuration, and the resulting error (MSE). Some of the tests conducted in previous studies obtained information that the Levenberg-Marquardt training algorithm has better performance than other algorithms in the backpropagation network, which produces the smallest average error with a test level of α = 5% which used 10 neurons in a hidden layer. The number of neurons in hidden layers varies depending on the number of neurons in the input layer. In this study an analysis of the performance of the Levenberg-Marquardt training algorithm was carried out with 5 neurons in the input layer, a number of n neurons in hidden layers (n = 2, 4, 5, 7, 9), and 1 neuron in the output layer. Performance analysis is based on network-generated errors. This study uses a mixed method, namely development research with quantitative and qualitative testing using ANOVA statistical tests. Based on the analysis, the Levenberg-Marquardt training algorithm produces the smallest error of 0.00014 ± 0.00018 on 9 neurons in hidden layers with lr = 0.5. Keywords: hidden layer, backpropogation, MSE, learning rate, Levenberg-Marquardt.