In Automatic Speech Recognition (ASR) the non-linear data projection provided by a one hidden layer Multilayer Perceptron (MLP), trained to recognize phonemes, and has previous experiments to provide feature enhancement substantially increased ASR performance, especially in noise. Previous attempts to apply an analogous approach to speaker identification have not succeeded in improving performance, except by combining MLP processed features with other features. We present test results for the TIMIT database which show that the advantage of MLP preprocessing for open set speaker identification increases with the number of speakers used to train the MLP and that improved identification is obtained as this number increases beyond sixty. We also present a method for selecting the speakers used for MLP training which further improves identification performance.
The esterification of oleic acid with 2-ethylhexanol in presence of sulfuric acid as homogeneous catalyst was investigated in this work to produce 2-ethylhexyl oleate (biodiesel) by using semi batch reactive distillation. The effect of reaction temperature (100 to 130°C), 2-ethylhexanol:oleic acid molar ratio (1:1 to 1:3) and catalysts concentration (0.2 to 1wt%) were studied. Higher conversion of 97% was achieved with operating conditions of reaction temperature of 130°C, molar ratio of free fatty acid to alcohol of 1:2 and catalyst concentration of 1wt%. A simulation was adopted from basic principles of the reactive distillation using MATLAB to describe the process. Good agreement was achieved.
The background subtraction is a leading technique adopted for detecting the moving objects in video surveillance systems. Various background subtraction models have been applied to tackle different challenges in many surveillance environments. In this paper, we propose a model of pixel-based color-histogram and Fuzzy C-means (FCM) to obtain the background model using cosine similarity (CS) to measure the closeness between the current pixel and the background model and eventually determine the background and foreground pixel according to a tuned threshold. The performance of this model is benchmarked on CDnet2014 dynamic scenes dataset using statistical metrics. The results show a better performance against the state-of the art
... Show MoreThis paper investigated the treatment of textile wastewater polluted with aniline blue (AB) by electrocoagulation process using stainless steel mesh electrodes with a horizontal arrangement. The experimental design involved the application of the response surface methodology (RSM) to find the mathematical model, by adjusting the current density (4-20 mA/cm2), distance between electrodes (0.5-3 cm), salt concentration (50-600 mg/l), initial dye concentration (50-250 mg/l), pH value (2-12 ) and experimental time (5-20 min). The results showed that time is the most important parameter affecting the performance of the electrocoagulation system. Maximum removal efficiency (96 %) was obtained at a current density of 20 mA/cm2, distance be
... Show MoreFree Radical Copolymerization of Styrene/ Methyl Methacrylate were prepared chemically under Nitrogen ,which was investigated, in the present of Benzoyl Peroxide as Initiator at concentration of 2 × 10-3 molar at 70 °C, which was carried out in Benzene as solvent to a certain low conversion . FT-IR spectra were used for determining of the monomer reactivity ratios ,which was obtained by employing the conventional linearization method of Fineman-Ross (F-R) and Kelen-Tüdos (K- T). The experimental results showed the average value for the Styrene r1 / Methyl Methacrylate r2 system, Sty r1 = 0.45 , MMA r2 = 0.38 in the (F–R) Method and r1 = 0.49 , r2 = 0.35 in the (K–T) Method, The Results of this indicated show the random distri
... Show MoreSupport vector machines (SVMs) are supervised learning models that analyze data for classification or regression. For classification, SVM is widely used by selecting an optimal hyperplane that separates two classes. SVM has very good accuracy and extremally robust comparing with some other classification methods such as logistics linear regression, random forest, k-nearest neighbor and naïve model. However, working with large datasets can cause many problems such as time-consuming and inefficient results. In this paper, the SVM has been modified by using a stochastic Gradient descent process. The modified method, stochastic gradient descent SVM (SGD-SVM), checked by using two simulation datasets. Since the classification of different ca
... Show MoreWe have investigated in this research, the contents of the electronic cigarette (Viber) and the emergence of the phenomenon of electronic smoking (vibing) were discussed, although the topic of smoking is one of the oldest topics on which many articles and research have been conducted, electronic smoking has not been studied according to statistical scientific research, we tried in this research to identify the concept of electronic smoking to sample the studied data and to deal with it in a scientific way. This research included conducting a statistical analysis using the factor analysis of a sample taken randomly from some colleges in Bab Al-medium in Baghdad with a size of (70) views where (КМО) and a (bartlett) tests
... Show MoreCoronavirus disease (COVID-19) is an acute disease that affects the respiratory system which initially appeared in Wuhan, China. In Feb 2019 the sickness began to spread swiftly throughout the entire planet, causing significant health, social, and economic problems. Time series is an important statistical method used to study and analyze a particular phenomenon, identify its pattern and factors, and use it to predict future values. The main focus of the research is to shed light on the study of SARIMA, NARNN, and hybrid models, expecting that the series comprises both linear and non-linear compounds, and that the ARIMA model can deal with the linear component and the NARNN model can deal with the non-linear component. The models
... Show More