The economic downturn in recent years has had a significant negative impact on corporates performance. In the last two years, as in the last years of 2010s, many companies have been influenced by the economic conditions and some have gone bankrupt. This has led to an increase in companies' financial risk. One of the significant branches of financial risk is the emph{company's credit risk}. Lenders and investors attach great importance to determining a company's credit risk when granting a credit facility. Credit risk means the possibility of default on repayment of facilities received by a company. There are various models for assessing credit risk using statistical models or machine learning. In this paper, we will investigate the machine learning task of the binary classification of firms into bankrupt and healthy based on the emph{spectral graph theory}. We first construct an emph{adjacency graph} from a list of firms with their corresponding emph{feature vectors}. Next, we first embed this graph into a one-dimensional Euclidean space and then into a two dimensional Euclidean space to obtain two lower-dimensional representations of the original data points. Finally, we apply the emph{support vector machine} and the emph{multi-layer perceptron} neural network techniques to proceed binary emph{node classification}. The results of the proposed method on the given dataset (selected firms of Tehran stock exchange market) show a comparative advantage over PCA method of emph{dimension reduction}. Finally, we conclude the paper with some discussions on further research directions.

Soil is a very complex phenomenon that includes organic materials, minerals, water, and air. The distribution of organic matter in the soil has a profound effect on biological activity, nutrient availability, soil and soil seed structure, and water holding capacity, and soil management in general. In this research, the relation between soil spectral reflectance using the Landsat 8 satellite data as well as the SRTM Elevation data and soil organic carbon has been investigated. In the proposed method, spectral reflection of data in the main bands of the Landsat 8 satellite is investigated and processed. In addition to the main bands, vegetation and lighting indices, and topographic features have been studied. In this study, a method for selecting effective indexes in increasing the accuracy of soil organic carbon modeling is presented. For this purpose, in the first step of modeling, Linear regression, Support Vector Machine regression, and Neural Network methods have been used for the connection between remote sensing data and soil organic carbon. To implement the proposed method, 100 soil samples in East Azerbaijan province have been used. According to RMSE and R2 statistical indices, which are the basis for evaluating the models, the neural network model was selected as the final model, and with the values of RMSE = 0.404, R2= 0.254, and RRMSE=46.597 is more accurate than the regression method. Due to the importance of dimensionality to increase accuracy and reduce the complexity of calculations, a genetic algorithm was proposed in this study. This efficient algorithm increases the accuracy of soil organic carbon modeling and eliminates additional indicators. After applying the genetic algorithm (GA) to the neural network model, we were able to achieve better accuracy, and the values of the baseline statistical indices were changed to RMSE = 0.279, R2 = 0.718, and RRMSE=27.116. Also, to check the efficiency of the genetic algorithm, the PCA algorithm was also implemented on the data and the comparison results showed that the genetic algorithm was successful in reducing dimensions along with increasing accuracy.

The search for effective treatments for complex diseases, while minimizing toxicity and side effects, has become crucial. However, identifying synergistic combinations of drugs is often a time-consuming and expensive process, relying on trial and error due to the vast search space involved. Addressing this issue, we present a deep learning framework in this study. Our framework utilizes a diverse set of features, including chemical structure, biomedical literature embedding, and biological network interaction data, to predict potential synergistic combinations. Additionally, we employ autoencoders and principal component analysis (PCA) for dimension reduction in sparse data. Through 10-fold cross-validation, we achieved an impressive 98 percent area under the curve (AUC), surpassing the performance of seven previous state-of-the-art approaches by an average of 8%.

In this paper, a mixed modeling approach for orthotropic laminated plates is developed. By adopting Hellinger-Reissner functional and dimension reduction method along the thickness, the governing equations were derived. By considering other theories i. e. classical plate theory, first order shear deformation theory and elasticity theory, the advantages of the current work are illustrated with some numerical results. Excellent agreements were observed by comparing the obtained results with three-dimensional elasticity theory for laminated thick plates. In the presented method, shear correction factor was not required for considering shear strain components. Furthermore, finite element simulation was implemented in Abaqus software by using two-dimensional shell elements and compared with obtained results. It is seen that although finite element model predicts good results for displacement field but it cannot provide any suitable results in thickness direction.

Introduction: Autism Spectrum Disorder (ASD) is a mental disorder and affects a person’, s linguis tic skills and social interactions. With the production of Functional Magnetic Resonance Imaging (fMRI) and the development of their processing tools, the use of these images in identifying and evaluating the brain function of autis tic people received a lot of attention. However, in this approach using the functional connectivity matrices leads to the creation of feature space with very high dimensions. Some of these features are dependent, unnecessary and additional, which reduces the quality of detection and increases the number of calculations. Therefore, regarding the large dimensions of the search space, the Particle Swarm Optimization (PSO) algorithm has been used as one of the powerful meta-heuris tic search tools in selecting the optimal features. Materials and Methods: To evaluate the capability of the proposed method, the principal component analysis (PCA) algorithm is used as a s tandard dimension reduction method. In this s tudy, the Support Vector Machines (SVM) classifier was used to detect autis tic and healthy persons on the ABIDE database data. Feature space has been generated based on a functional connectivity matrix which has 6670 dimensions. Results: SVM accuracy in high-dimensional specialty space is 56%. The proposed method based on PSO eliminates 3442 redundant features and increases classification accuracy up to 62. 19 % that performs better than PCA. The findings show that this meta-heuris tic algorithm by removing almos t half of the features results in a 6% increase in classification precision. Conclusion: The results indicate the ability of SVM in comparison with the Random Forest and K-Neares t Neighbor (KNN). PSO algorithm was used for dimension reduction of the input data space.

These days, the most important areas of research in many different applications, with different tools, are focused on how to get awareness. One of the serious applications is the awareness of the behavior and activities of patients. The importance is due to the need of ubiquitous medical care for individuals. That the doctor knows the patient's physical condition, sometimes is very important. Of course, there are other important applications for this information. There are a variety of methods and tools for measurement, gathering, and analysis of the physical behaviors and activities’ information. One of the most successful tools for this aim are ubiquitous intelligent electronic devices, specifically smartphones, and smart watches. There are many sensors in these devices, some of which can be used to understand the activities of daily living. As an output result, these sensors produce many raw data. Thus, it is needed to process these information and recognize the individual behavior of the output of this processing. In this paper, the basic components of the analysis phase for this process have been proposed. Simulations validate the benefits and superiority of this method.

Nowadays, users can share their ideas and opinions with widespread access to the Internet and especially social networks. On the other hand, the analysis of people's feelings and ideas can play a significant role in the decision making of organizations and producers. Hence, sentiment analysis or opinion mining is an important field in natural language processing. One of the most common ways to solve such problems is machine learning methods, which creates a model for mapping features to the desired output. One challenge of using machine learning methods in NLP fields is feature selection and extraction among a large number of early features to achieve models with high accuracy. In fact, the high number of features not only cause computational and temporal problems but also have undesirable effects on model accuracy. Studies show that different methods have been used for feature extraction or selection. Some of these methods are based on selecting important features from feature sets such as Principal Component Analysis (PCA) based methods. Some other methods map original features to new ones with less dimensions but with the same semantic relations like neural networks. For example, sparse feature vectors can be converted to dense embedding vectors using neural network-based methods. Some others use feature set clustering methods and extract less dimension features set like NMF based methods. In this paper, we compare the performance of three methods from these different classes in different dataset sizes. In this study, we use two compression methods using Singular Value Decomposition (SVD) that is based on selecting more important attributes and non-Negative Matrix Factorization (NMF) that is based on clustering early features and one Auto-Encoder based method which convert early features to new feature set with the same semantic relations. We compare these methods performance in extracting more effective and fewer features on sentiment analysis task in the Persian dataset. Also, the impact of the compression level and dataset size on the accuracy of the model has been evaluated. Studies show that compression not only reduces computational and time costs but can also increase the accuracy of the model. For experimental analysis, we use the Sentipers dataset that contains more than 19000 samples of user opinions about digital products and sample representation is done with bag-of-words vectors. The size of bag-of-words vectors or feature vectors is very large because it is the same as vocabulary size. We set up our experiment with 4 sub-datasets with different sizes and show the effect of different compression performance on various compression levels (feature count) based on the size of dataset size. According to experiment results of classification with SVM, feature compression using the neural network from 7700 to 2000 features not only increases the speed of processing and reduces storage costs but also increases the accuracy of the model from 77. 05% to 77. 85% in the largest dataset contains about 19000 samples. Also in the small dataset, the SVD approach can generate better results and by 2000 features from 7700 original features can obtain 63. 92 % accuracy compared to 63. 57 % early accuracy. Furthermore, the results indicate that compression based on neural network in large dataset with low dimension feature sets is much better than other approaches, so that with only 100 features extracted by neural network-based auto-encoder, the system achieves acceptable 74. 46% accuracy against SVD accuracy 67. 15% and NMF accuracy 64. 09% and the base model accuracy 77. 05% with 7700 features.