Hidden Markov Model is a popular statisical method that is used in continious and DISCRETE speech recognition. The probability DENSITY function of observation vectors in each state is estimated with DISCRETE DENSITY or continious DENSITY modeling. The performance (in correct word recognition rate) of continious DENSITY is higher than DISCRETE DENSITY HMM, but its computation complexity is very high, especially in very large DISCRETE utterance recognition problems. For real time implementation of very large DISCRETE utterance recognition, we must use DISCRETE DENSITY HMM (DDHMM). To increase the performance of DDHMM, one usual solution is fuzzy interpolation. In this study, we present a new method named Gaussian interpolation. We implemented and compared the performance of two types of interpolation methods for 1500 Persian speech command words. Results show that precision and flexibility of Gaussian interpolation is better thanthose of the fuzzy interpolation.

The effectiveness of hydraulic fracturing fluid injection is influenced by numerous factors, including pre-existing discontinuities such as DISCRETE fracture networks (DFNs). Among the geometric characteristics of DFNs, fracture DENSITY is a critical factor. In deep reservoirs, which often consist of hot dry rock (HDR), thermal conduction through the rock and fluid, as well as advection and convective heat transfer within the fluid, can significantly impact fluid–rock interactions. This study examines the influence of DFN DENSITY on hydraulic fracture (HF) propagation in HDR, with a particular focus on the thermo-hydro-mechanical (THM) behavior of HDR using the combined finite-DISCRETE element method (FDEM). Key controlling factors, such as flow rate, fluid kinematic viscosity, in-situ stress magnitude, pre-existing fracture aperture, and working fluid temperature, are analyzed. The findings highlight the significant role of DFN DENSITY in determining the pattern and extent of HF propagation under varying conditions. Additionally, the interaction between the working fluid and DFNs is shown to vary considerably with changes in these controlling factors. However, the study reveals that variations in DFN DENSITY or the values of the controlling factors have minimal impact on the temperature field. This is attributed to the rapid heat exchange between the cold fluid and the HDR, which quickly raises the fluid temperature, resulting in negligible temperature variations.

Distance-based clustering methods categorize samples by optimizing a global criterion, finding ellipsoid clusters with roughly equal sizes. In contrast, DENSITY-based clustering techniques form clusters with arbitrary shapes and sizes by optimizing a local criterion. Most of these methods have several hyper-parameters, and their performance is highly dependent on the hyper-parameter setup. Recently, a Gaussian DENSITY Distance (GDD) approach was proposed to optimize local criteria in terms of distance and DENSITY properties of samples. GDD can find clusters with different shapes and sizes without any free parameters. However, it may fail to discover the appropriate clusters due to the interfering of clustered samples in estimating the DENSITY and distance properties of remaining unclustered samples. Here, we introduce Adaptive GDD (AGDD), which eliminates the inappropriate effect of clustered samples by adaptively updating the parameters during clustering. It is stable and can identify clusters with various shapes, sizes, and densities without adding extra parameters. The distance metrics calculating the dissimilarity between samples can affect the clustering performance. The effect of different distance measurements is also analyzed on the method. The experimental results conducted on several well-known datasets show the effectiveness of the proposed AGDD method compared to the other well-known clustering methods.

Development of bacterial databases is crucial and every year the number of prokaryotic genome is increasing. The problem of identifying genes in genomic DNA sequences by computational methods has attracted considerable research attention in recent years.A Full automatic and self-train Gene finder is presented in this research. This system uses non-looped HMM to measure of statistical significance for Genes in prokaryotic genomes. Design of this software was done in three main programs and developed in C++. First program is presented for extraction the DATA (Long non-overlapping ORFs) to train the machine learning algorithm in a self-training method. Second program is related to the training stage. In this stage, HMM is trained with the data that obtained in the previous stage. We model standard 'text book genes' with an unbroken open reading frame. In the last program, The Long ORFs is scored with the trained system. Finally, Genes are selected on the base on their lengths and scores. Our Gene finder can predicts genes with Specifity >96 and Sensetivity>84. The result shows that overall performance of our software matches other methods that are designed by others.

In this paper, a new Hidden Markov Model (HMM)-based face recognition system is proposed. As a novel point despite of five-state HMM used in pervious researches, we used 7-state HMM to cover more details. Indeed we add two new face regions, eyebrows and chin, to the model. As another novel point, we used a small number of quantized Singular Values Decomposition (SVD) coefficients as features describing blocks of face images. This makes the system very fast. The system has been evaluated on the Olivetti Research Laboratory (ORL) face database. In order to additional reduction in computational complexity and memory consumption the images are resized to 64x64 jpeg format. Before anything, an order-statistic filter is used as a preprocessing operation. Then a top-down sequence of overlapping sub-image blocks is considered. Using quantized SVD coefficients of these blocks, each face is considered as a numerical sequence that can be easily modeled by HMM. The system has been examined on 400 face images of the Olivetti Research Laboratory (ORL) face database. The experiments showed a recognition rate of 99%, using half of the images for training. The system has been evaluated on 64x64 jpeg resized YALE database too. This database contains 165 face images with 231x195 pgm format. Using five training image, we obtained 97.78% recognition rate where for six training images the recognition rate was 100%, a record in the literature. The proposed method is compared with the best researches in the literature. The results show that the proposed method is the fastest one, having approximately 100%recognition rate.

Intrusion Detection Systems (IDSs) are security tools widely used in computer networks. While they seem to be promising technologies, they pose some serious drawbacks: When utilized in large and high traffic networks, IDSs generate high volumes of low-level alerts which are hardly manageable. Accordingly, there emerged a recent track of security research, focused on alert correlation, which extracts useful and high-level alerts, and helps to make timely decisions when a security breach occurs.In this paper, we propose an alert correlation system consisting of two major components; first, we introduce an Attack Scenario Extraction Algorithm (ASEA), which mines the stream of alerts for attack scenarios. The ASEA has a relatively good performance, both in speed and memory consumption. Contrary to previous approaches, the ASEA combines both prior knowledge as well as statistical relationships. Second, we propose a Hidden Markov Model (HMM)-based correlation method of intrusion alerts, red from different IDS sensors across an enterprise. We use HMM to predict the next attack class of the intruder, also known as plan recognition. This component has two advantages: Firstly, it does not require any usage or modeling of network topology, system vulnerabilities, and system configurations; Secondly, as we perform high-level prediction, the model is more robust against over- fitting. In contrast, other published plan-recognition methods try to predict exactly the next attacker action.We applied our system to DARPA 2000 intrusion detection scenario dataset.The ASEA experiment shows that it can extract attack strategies efficiently.We evaluated our plan-recognition component both with supervised and unsupervised learning techniques using DARPA 2000 dataset. To the best of our knowledge, this is the first unsupervised method in attack-plan recognition.

Membranes are widely used in various engineering applications such as the design stage of microphones, pumps, pressure regulators, and other acoustical applications. This paper investigates the numerical aspects for free vibration analysis of homogeneous and non-homogeneous rectangular and square membranes. The method of DISCRETE singular convolution is employed. The results are obtained for different DENSITY case and aspect ratios. Numerical results are presented and compared with that available in the literature. The results show that the regularized Shannon delta kernel based DISCRETE singular convolution algorithm produces accurate frequency values.