Dramatic changes in digital communication and exchange of image, audio, video and text files result in a suitable field for interpersonal transfers of hidden information. Therefore, nowadays, preserving channel security and intellectual property and access to hidden information make new fields of researches naming steganography, watermarking and steganalysis. Steganalysis as a binary classification distinguish clean signals from stego signals. Features extracted from time and transform domain are proper for this classifier. Some of steganalysis methods are depended on a specific steganography algorithm and others are independent. The second group of methods are called Universal steganalysis. Universal steganalysis methods are widely used in applications because of their independency to steganography algorithms. These algorithms are based on characteristics such as distortion measurements, higher order statistics and other similar features. In this research we try to achieve more reliable and accurate results using analytical review of features, choose more effective of them and optimize SVM performance. In new researches Mel Frequency Cepstral Coefficient and Markov transition probability matrix coefficients are used to steganalysis design. In this paper we consider two facts. First, MFCC extract signal features in transform domain similar to human hearing model, which is more sensitive to low frequency signals. As a result, in this method there is more hidden information mostly in higher frequency audio signals. Therefore, it is suggested to use reversed MFCC. Second, there is an interframe correlation in audio signals which is useful as an information hiding effect. For the first time, in this research, this features is used in steganalysis field. To have more accurate and stable results, we use recursive feature elimination with correlation bias reduction for SVM. To implement suggested algorithm, we use two different data sets from TIMIT and GRID. For each data sets, Steghide and LSB-Matching steganography methods implement with 20 and 50 percent capacity. In addition, one of the LIBSVM 3. 2 toolboxes is sued for implementation. Finally, the results show accuracy of steganalysis, four to six percent increase in comparison with previous methods. The ROC of methods clearly shows this improvement.

This research used modern machine learning ways to predict the stages of primary biliary cholangitis using data from the Mayo Clinic trial. The research aims to obtain high prediction accuracy while representing balanced evaluation metrics. Important techniques include automated hyperparameters optimization with Optuna and Recursive Feature Elimination to improve model performance. Pre-processing included handling missing values, encoding of categorical features, and addressing class imbalances using SMOTE. A total of twelve machine learning algorithms are evaluated with ensemble-based models such as CatBoost and Extra Trees producing much better results. Evaluation metrics take into account all model predictions, including accuracy, precision, recall, F1 score, and ROC-AUC for performing balanced and interpretative evaluations of performances critical for imbalanced datasets. This endeavor includes clinical and laboratory information illustrating the prospect of machine learning in advancing therapeutic diagnosis, emphasizing the rigor and robustness in evaluation laid groundwork for future research to encompass even more generalizable and robust diagnostic tools.

The price of EU Allowances (EUAs) serves as a mechanism for managing greenhouse gas emissions, influenced by various factors such as economic, financial, political, and other indicators. This study assesses the impact of 31 different energy, financial, and commodity indices on EUA prices. To achieve this, a hybrid data-driven approach is employed. Initially, the Autoregressive Integrated Moving Average (ARIMA) algorithm cleans the relevant data. Subsequently, the Variational Mode Decomposition (VMD) method decomposes the indices into intrinsic mode functions (IMFs). These IMFs are categorized into three-time scales: short-term, medium-term, and long-term. Next, by integrating Recursive Feature Elimination (RFE) and Random Forest (RF) with cross-validation, the most influential features across these time scales are selected. The results demonstrate that the model achieves higher accuracy in medium-term and long-term time scales. Forecasting the price fluctuations of EUAs using these hybrid approaches can contribute to more precise and timely decision-making in policy formulation and investment strategies.

Our main objective is to study some qualitative behavior of the solutions of the difference equation xn+1 = gxn−k + (axn + bxn−k) / (cxn − dxn−k), n = 0, 1, 2, ..., where the initial conditions x−k,..., x−1, x0 are arbitrary positive real numbers and the coefficients, a, b, c and d are positive constants, while k is a positive integer number.

Introduction: P300 speller is a kind of Brain-Computer Interface (BCI) system in which the user may type words by using the responses obtained from human focus on different characters. The high sensitivity of brain signals against noise in parallel with the similarity of responses obtained from the user focus on different characters makes it difficult to classify the characters based on their respective P300 wave. On the other hand, all areas of the brain does not carry useful P300 information. Methods: In this study, a new method is proposed to improve the performance of speller system which is based on selecting optimal P300 channels. In the proposed method, recursive elimination algorithm is presented for channel optimization, which utilizes deep learning concept (e. g. Convolutional Neural Network) as its cost function. The proposed method is examined on a data set from EEG signals recorded in a P300 speller system, including 64 different channels of responses to 29 characters. Then, its performance is compared with some existing methods. Results: The obtained results showed the ability of the proposed method in recognizing the characters in such way that it could accurately (i. e. 97. 34%) detect 29 characters by using only 24 out of all 64 electrodes. Conclusion: Applying the proposed method in speller systems led to considerable improvement in classification of characters compared to its alternatives. Several experiments proved that utilizing the proposed scheme may increases the accuracy almost 12. 9 percent compared to non-optimized case in parallel with reduction of the number of involved channels by approximately 1/3. Based on these results, the proposed method may be considered as an effective choice for application in P300 speller systems, thanks to reduction of the complexity of the system which is caused by the reduced number of channels and, on the other hand, due to its potential in increasing the accuracy of character recognition.

Multi-label classification aims at assigning more than one label to each instance. Many real-world multi-label classification tasks are high dimensional, leading to reduced performance of traditional classifiers. Feature selection is a common approach to tackle this issue by choosing prominent features. Multi-label feature selection is an NP-hard approach, and so far, some swarm intelligence-based strategies and have been proposed to find a near optimal solution within a reasonable time. In this paper, a hybrid intelligence algorithm based on the binary algorithm of particle swarm optimization and a novel local search strategy has been proposed to select a set of prominent features. To this aim, features are divided into two categories based on the extension rate and the relationship between the output and the local search strategy to increase the convergence speed. The first group features have more similarity to class and less similarity to other features, and the second is redundant and less relevant features. Accordingly, a local operator is added to the particle swarm optimization algorithm to reduce redundant features and keep relevant ones among each solution. The aim of this operator leads to enhance the convergence speed of the proposed algorithm compared to other algorithms presented in this field. Evaluation of the proposed solution and the proposed statistical test shows that the proposed approach improves different classification criteria of multi-label classification and outperforms other methods in most cases. Also in cases where achieving higher accuracy is more important than time, it is more appropriate to use this method.

This study aimed to investigate the spatial changes of soil salinity using RF model in a part of Eyvanekey Plain (Semnan Province 2018). Grid sampling with 100 m intervals (106 samples) was taken from 105 ha of soils developed on marl and gravely alluviums. The land uses were pistachio plantations with furrow irrigation and abandoned land. The maximum EC was (173.2 and 34 dS/m) in the abandoned and furrow irrigation pistachio plantations respectively. The main factors of salinization were saline marls, saline irrigation water, and high PET. The R2 for the salinity prediction map by RF model was 0.49, and the most important covariates were normalized difference salinity index (NDSI), topographic wetness index (TWI), Channel Network Base Level (CNBL), normalized difference vegetation index (NDVI), and modified soil vegetation index (SAVI). Spectral ratio indices derived from Landsat 8 contributed the most to the soil salinity prediction. Out of 5 main auxiliary variables, 3 variables are related to spectral ratio indices and the reason was the presence of salt on the soil in the studied area. Using NDVI with other salinity and moisture indices improved the salinity prediction model. Examining the results of covariates correlation and the implementation of recursive feature elimination showed that many covariates increase model complexity and prediction error. Recursive feature elimination helped to simplify the model by identifying the most important covariates. The salinity prediction map by random forest was consistent with the field observations and clearly defined the critical saline area.

In this paper it has been attempted to investigate the capability of the consumption-based capital asset pricing model (CCAPM), using the general method of moment (GMM), with regard to the Epstien-zin recursive preferences model for Iran's capital market. Generally speaking, recursive utility permits disentangling of the two psychologically separate concepts of risk aversion and elasticity of intertemporal substitution which are constrained to be equal to the inverse of each other for the traditional time-additive utility functions. Rather than using the stock market as a proxy for wealth, we constructed a more comprehensive return which is the weighted average of stock index return, labor wage growth (as a proxy for human capital return), housing return and deposit return. The empirical results demonstrate that the signs of the coefficient of the relative risk aversion and the intertemporal elasticity of substitution are the same, which means that investors have homogeneous attitudes toward the risk across the states of nature and the risk over time in Iran but different ones in their values. Therefore, the assumption that the relative risk aversion is equal to the reciprocal of the elasticity of substitution is not valid in Iran's stock market.