In order to predict the behavior of soil-related phenomena, it is necessary to have knowledge about unsaturated flow and using models that provide optimal estimates of the retention curve and hydraulic conductivity of soils. Despite the widespread use of the classic van Genuchten-Mualem model (VGM), this model usually performs poorly in predicting hydraulic conductivity and modification of some of its parameters seems necessary. In this research, 283 soils from different textures of the UNSODA bank were selected and divided into two sections of calibration and validation and their soil parameters were exported and categorized. Then, by defining the modified unsaturated hydraulic conductivity (Ksc) instead of the saturated hydraulic conductivity (Ks) and determining the limits for l and n parameters, the hydraulic conductivity-moisture function of VGM were solved using 24600 pairs of points li and nj for each soil of the three main soil texture classes. In the following, the optimal l value (l̂) of each texture class was selected based on the minimum value of the hydraulic conductivity estimation error using the root mean square error (RMSE) index and the n values that had created the minimum errors, were selected as the optimal pore size distribution coefficients of the hydraulic conductivity-moisture function (n̂opt). In order to create pedotransfer functions for estimating n̂opt, we ran stepwise regression in MATLAB software considering the condition of statistical significance (P-value=0.05) for independent variables and functions for each soil texture class. After creating pedotransfer functions, the results of the proposed method of this research (MVGM) were compared with the VGM results using RMSE and Nash-Sutcliffe (NSE) indices. The results showed that in both sections of creation and validation functions, the MVGM performed better in estimating hydraulic conductivity and had a higher efficiency index in all textural classes of soil.

Kochia (Kochia scoparia L. Schard) is an annual, halophyte and drought resistant plant, that it can be irrigated with saline water and a valuable source for forage under drought and saline ecosystems. In order to evaluate germination characteristics of kochia, an experiment was conducted during 2009 at Physiology laboratory of Ferdowsi University of Mashhad, Iran. This experiment was conducted in a completely randomized design with four replications. Germination was evaluated at 5, 10, 15, 20, 25, 30, 35 and 40oC under dark germinator with 50-60 percentage relative humidity. The results showed that the highest germination percentage was obtained at 20-30oC and the lowest obtained at 40oC. The longest and the shortest period to 20 and 50 germination percentages were recorded to 5-10oC and 20-30oC, respectively. The longest and the shortest period to 80 percentage germination were belonging to 15 and 30oC, respectively. Based on Five parameters Beta model, base, optimum and ceiling temperatures for kochia estimated 3.4, 25 and 43.3oC, respectively. However, seed of this plant is able to germinate in wide temperature range.

Distributed Denial of Service (DDoS) attacks are among the primary concerns in internet security today. Machine learning can be exploited to detect such attacks. In this paper, a multi-layer perceptron model is proposed and implemented using deep machine learning to distinguish between malicious and normal traffic based on their behavioral patterns. The proposed model is trained and tested using the CICDDoS2019 dataset. To remove irrelevant and redundant data from the dataset and increase learning accuracy, feature selection is used to select and extract the most effective features that allow us to detect these attacks. Moreover, we use the grid search algorithm to acquire optimum values of the model’s hyperparameters among the parameters’ space. In addition, the sensitivity of accuracy of the model to variations of an input parameter is analyzed. Finally, the effectiveness of the presented model is validated in comparison with some state-of-the-art works.

In traditional speech processing, feature extraction and classification were conducted as separate steps. The advent of deep neural networks has enabled methods that simultaneously model the relationship between acoustic and phonetic characteristics of speech while classifying it directly from the raw waveform. The first convolutional layer in these networks acts as a filter bank. To enhance interpretability and reduce the number of parameters, researchers have explored the use of parametric filters, with the SincNet architecture being a notable advancement. In SincNet's initial convolutional layer, rectangular bandpass filters are learned instead of fully trainable filters. This approach allows for modeling with fewer parameters, thereby improving the network's convergence speed and accuracy. Analyzing the learned filter bank also provides valuable insights into the model's performance. The reduction in parameters, along with increased accuracy and interpretability, has led to the adoption of various parametric filters and deep architectures across diverse speech processing applications. This paper introduces different types of parametric filters and discusses their integration into various deep architectures. Additionally, it examines the specific applications in speech processing where these filters have proven effective.

In many applications we have to encountered with bounded dependent variables. Beta regression model can be used to deal with these kinds of response variables. In this paper we aim to study spatially correlated responses in the unit interval. Initially we introduce spatial Beta generalized linear mixed model in which the spatial correlation is captured through a random effect. Then the performances of the proposed model is evaluated via a simulation study, implementing Bayesian approach for parameter estimation.Finally the application of this model on two real data sets about migration rate and divorce rate in Iran are presented.

Facilities built in areas affected by earthquake activity, such as tunnels, which have always been an integral part of human life, must withstand both dynamic and static loading. It has led to the need for practical studies on the effects of earthquakes on underground structures and the factors affecting their destruction. For this purpose, in this research, at first different patterns of tunnel’s excavation were investigated and by using Plaxis 2D software and based on Tabas earthquake in Iran, sensitivity analysis on geotechnical parameters of the soil surrounding tunnel such as cohesion, friction angle, unit weight and modulus of elasticity was carried out, and the parameters whose changes have the greatest and least effects on the bending moment changes on the tunnel lining are introduced. The results show that tunnel excavation patterns significantly affect the bending moment, axial forces, displacements, and surface settlement of the tunnel. Often, by dividing tunnel excavation area to small parts, the values of bending moment, axial forces, displacements, and surface settlement of the tunnel decreases in static analysis. Also, outputs of sensitivity analysis on geotechnical parameters of the soil surrounding tunnel showed that modulus of elasticity of the soil surrounding tunnel has the most effect and cohesion changes have the least effect on bending moment induced on tunnel lining..

When evaluating companies and investment plans, most analysts use a discount rate that is derived from CAPM models. The Beta in these models usually represent risks and opportunities of the relative industry, with almost no attention to the risks that are already included in the Beta. This ignorance in risk measurement could ultimately impair shareholders value. What makes things critical is that by adjusting risks and opportunities in Beta, the result of investment plans and company valuation could be much different. In this paper we use 1 to 10 years of monthly return data for all industries of Tehran Stock Exchange and Iran Fara Bourse and suggest an adjusted Beta for each industry which is stripped of the dazzling effects of the debts and growth opportunities. We separately account for breaking down Beta into Beta of growth opportunities and Beta of existing assets for each industry in various timelines between 1 to 10 years. Our results showed that the Beta of growth opportunities is bigger than the Beta of assets for almost all industries. The mentioned Betas can make a big difference in cost of capital and we suggest using them when evaluating investment plans, development plans, valuation of companies and even start-ups.