Minimal neural network topology optimization for aesthetic classification

On predicting annual output energy of 4-terminal perovskite/silicon tandem PV cells for building integrated photovoltaic application using machine learning

Building integrated photovoltaic (BIPV), based on tandem PV cells, is considered a new alternative for combining solar energy with buildings. Accurately predicting the BIPV-harvested annual output energy (E o u t , a n n u a l ) is crucial for evaluating the BIPV performance. Machine learning (ML) is a potential candidate for solving such a problem without the time-consuming process of experimental investigations. This contribution proposes an artificial neural network (ANN) to predict the E o u t , a n n u a l of 4-terminal perovskite/silicon (psk/Si) PV cells under realistic environmental conditions. The input variables of the proposed model consist of the input solar irradiance (P i n ), incident light's angle (A i n ), the PV module's temperature (T m o d ), the psk absorber's thickness (T h p s k ), and the psk absorber's bandgap (B p s k ). The input data were received from the simulated results. This work also evaluates the degree of importance of each input variable and optimizes the architecture of the ANN using the surrogate algorithm before predictions. The optimized ANN-3 (three hidden layers) model shows superior performance indicators, including a mean squared error of MSE = 0.02283, correlation coefficient R = 0.99999, and Willmott's index of agreement I w = 0.99999. Consequently, the predicted highest E o u t , a n n u a l at B p s k of 1.71 eV is 297.73, 115.01, 193.98, and 97.6 kWh/m2 for the rooftop, east, south, and west facades, respectively.

Predicting mechanical properties of silk from its amino acid sequences via machine learning

The silk fiber is increasingly being sought for its superior mechanical properties, biocompatibility, and eco-friendliness, making it promising as a base material for various applications. One of the characteristics of protein fibers, such as silk, is that their mechanical properties are significantly dependent on the amino acid sequence. Numerous studies have been conducted to determine the specific relationship between the amino acid sequence of silk and its mechanical properties. Still, the relationship between the amino acid sequence of silk and its mechanical properties is yet to be clarified. Other fields have adopted machine learning (ML) to establish a relationship between the inputs, such as the ratio of different input material compositions and the resulting mechanical properties. We have proposed a method to convert the amino acid sequence into numerical values for input and succeeded in predicting the mechanical properties of silk from its amino acid sequences. Our study sheds light on predicting mechanical properties of silk fiber from respective amino acid sequences.

Validation of an Aesthetic Assessment System for Commercial Tasks

Automatic prediction of the aesthetic value of images has received increasing attention in recent years. This is due, on the one hand, to the potential impact that predicting the aesthetic value has on practical applications. Even so, it remains a difficult task given the subjectivity and complexity of the problem. An image aesthetics assessment system was developed in recent years by our research group. In this work, its potential to be applied in commercial tasks is tested. With this objective, a set of three portals and three real estate agencies in Spain were taken as case studies. Images of their websites were taken to build the experimental dataset and a validation method was developed to test their original order with another proposed one according to their aesthetic value. So, in this new order, the images that have the high aesthetic score by the AI system will occupy the first positions of the portal. Relevant results were obtained, with an average increase of 52.54% in the number of clicks on the ads, in the experiment with Real Estate portals. A statistical analysis prove that there is a significant difference in the number of clicks after selecting the images with the AI system.

Determining the number of hidden layer and hidden neuron of neural network for wind speed prediction

Artificial neural network (ANN) is one of the techniques in artificial intelligence, which has been widely applied in many fields for prediction purposes, including wind speed prediction. The aims of this research is to determine the topology of neural network that are used to predict wind speed. Topology determination means finding the hidden layers number and the hidden neurons number for corresponding hidden layer in the neural network. The difference between this research and previous research is that the objective function of this research is regression, while the objective function of previous research is classification. Determination of the topology of the neural network using principal component analysis (PCA) and K-means clustering. PCA is used to determine the hidden layers number, while clustering is used to determine the hidden neurons number for corresponding hidden layer. The selected topology is then used to predict wind speed. Then the performance of topology determination using PCA and clustering is then compared with several other methods. The results of the experiment show that the performance of the neural network topology determined using PCA and clustering has better performance than the other methods being compared. Performance is determined based on the RMSE value, the smaller the RMSE value, the better the neural network performance. In future research, it is necessary to apply a correlation or relationship between input attribute and output attribute and then analyzed, prior to conducting PCA and clustering analysis.