Journal of Artificial Intelligence and Data Mining
Volume:2 Issue: 1, Winter-Spring 2014

One of the most common positioning errors in panoramic radiography is palatoglossal air space above the apices of the root of maxillary teeth. It causes a radiolucency obscuring the apices of maxillary teeth. In the case of this positioning error, the imaging should be repeated. This causes the patient be exposed to radiation again. To avoid the repetition of exposing harmful X-rays to the patient, it is necessary to improve the panoramic images. This paper presents a new automatic panoramic image enhancement method to reduce the effect of this positioning error. Experimental results indicate that the enhanced panoramic images provide with adequate diagnostic information specially in maxilla sinusoid region. Hence, this technique dispenses the need for repetition of X-ray imaging.

Often in modeling the engineering optimization design problems, the value of objective function(s) is not clearly defined in terms of design variables. Instead it is obtained by some numerical analysis such as FE structural analysis, fluid mechanic analysis, and thermodynamic analysis, etc. Yet, the numerical analyses are considerably time consuming to obtain the final value of objective function(s). For the reason of reducing the number of analyses as few as possible our methodology works as a supporting tool to the meta-models. The research in meta-modeling for multiobjective optimization are relatively young and there is still much to do. Here is shown that visualizing the problem on the basis of the randomly sampled geometrical big-data of computer aided design (CAD) and computer aided engineering (CAE) simulation results, combined with utilizing classification tool of data mining could be effective as a supporting system to the available meta-modeling approaches.
To evaluate the effectiveness of the proposed method a study case in 3D wing optimal design is given. Along with the study case, it is discussed that how effective the proposed methodology could be in further practical engineering design problems.

Hand gesture recognition is very important to communicate in sign language. In this paper, an effective object tracking and hand gesture recognition method is proposed. This method is combination of two well-known approaches, the mean shift and the motion detection algorithm. The mean shift algorithm can track objects based on the color, then when hand passes the face occlusion happens. Several solutions such as the particle filter, kalman filter and dynamic programming tracking have been used, but they are complicated, time consuming and so expensive. The proposed method is so easy, fast, efficient and low cost. In the first step, the motion detection algorithm subtracts the previous frame from the current frame to obtain the changes between two images and white pixels (motion level) are detected by using the threshold level. Then the mean shift algorithm is applied for tracking the hand motion. Simulation results show this method is faster than two times to compared with the old common algorithms.

The enormous growth of the World Wide Web in recent years has made it necessary to perform resource discovery efficiently. For a crawler it is not an simple task to download the domain specific web pages. This unfocused approach often shows undesired results. Therefore, several new ideas have been proposed, among them a key technique is focused crawling which is able to crawl particular topical portions of the World Wide Web quickly without having to explore all web pages. Focused crawling is a technique which is able to crawled particular topics quickly and efficiently without exploring all WebPages. The proposed approach does not only use keywords for the crawl, but rely on high-level background knowledge with concepts and relations, which are compared with the texts of the searched page.
In this paper a combined crawling strategy is proposed that integrates the link analysis algorithm with association metric. An approach is followed to find out the relevant pages before the process of crawling and to prioritizing the URL queue for downloading higher relevant pages, to an optimal level based on domain dependent ontology. This strategy make use of ontology to estimate the semantic contents of the URL without exploring which in turn strengthen the ordering metric for URL queue and leads to the retrieval of most relevant pages.

Many researchers adopt Local Binary Pattern for pattern analysis. However, the long histogram created by Local Binary Pattern is not suitable for large-scale facial database. This paper presents a simple facial pattern descriptor for facial expression recognition. Local pattern is computed based on local gradient flow from one side to another side through the center pixel in a 3x3 pixels region. The center pixel of that region is represented by two separate two-bit binary patterns, named as Local Gradient Pattern-LGP for that pixel. LGP pattern is extracted from each pixel. Facial image is divided into 81 equal sized blocks and the histograms of local LGP features for all 81 blocks are concatenated to build the feature vector. Experimental results prove that the proposed technique along with Support Vector Machine is effective for facial expression recognition.

For multi-objective optimal reactive power dispatch (MORPD), a new approach is proposed where simultaneous minimization of the active power transmission loss, the bus voltage deviation and the voltage stability index of a power system are achieved. Optimal settings of continuous and discrete control variables (e.g. generator voltages, tap positions of tap changing transformers and the number of shunt reactive compensation devices to be switched)are determined. MORPD is solved using particle swarm optimization (PSO). Also, Pareto Optimality PSO (POPSO) is proposed to improve the performance of the multi-objective optimization task defined with competing and non-commensurable objectives. The decision maker requires managing a representative Pareto-optimal set which is being provided by imposition of a hierarchical clustering algorithm. The proposed approach was tested using IEEE 30-bus and IEEE 118-bus test systems. When simulation results are compared with several commonly used algorithms, they indicate better performance and good potential for their efficient applications in solving MORPD problems.

Speech Emotion Recognition (SER) is a new and challenging research area with a wide range of applications in man-machine interactions. The aim of a SER system is to recognize human emotion by analyzing the acoustics of speech sound. In this study, we propose Spectral Pattern features (SPs) and Harmonic Energy features (HEs) for emotion recognition. These features extracted from the spectrogram of speech signal using image processing techniques. For this purpose, details in the spectrogram image are firstly highlighted using histogram equalization technique. Then, directional filters are applied to decompose the image into 6 directional components. Finally, binary masking approach is employed to extract SPs from sub-banded images. The proposed HEs are also extracted by implementing the band pass filters on the spectrogram image. The extracted features are reduced in dimensions using a filtering feature selection algorithm based on fisher discriminant ratio. The classification accuracy of the pro-posed SER system has been evaluated using the 10-fold cross-validation technique on the Berlin database. The average recognition rate of 88.37% and 85.04% were achieved for females and males, respectively. By considering the total number of males and females samples, the overall recognition rate of 86.91% was obtained.

The record of human brain neural activities, namely electroencephalogram (EEG), is generally known as a non-stationary and nonlinear signal. In many applications, it is useful to divide the EEGs into segments within which the signals can be considered stationary. Combination of empirical mode decomposition (EMD) and Hilbert transform, called Hilbert-Huang transform (HHT), is a new and powerful tool in signal processing. Unlike traditional time-frequency approaches, HHT exploits the nonlinearity of the medium and non-stationarity of the EEG signals. In addition, we use singular spectrum analysis (SSA) in the pre-processing step as an effective noise removal approach. By using synthetic and real EEG signals, the proposed method is compared with wavelet generalized likelihood ratio (WGLR) as a well-known signal segmentation method. The simulation results indicate the performance superiority of the proposed method.

Yarn tenacity is one of the most important properties in yarn production. This paper addresses modeling of yarn tenacity as well as optimally determining the amounts of the effective inputs to produce yarn with desired tenacity. The artificial neural network is used as a suitable structure for tenacity modeling of cotton yarn with 30 Ne. As the first step for modeling, the empirical data is collected for cotton yarns. Then, the structure of the neural network is determined and its parameters are adjusted by back propagation method. The efficiency and accuracy of the neural model is measured based on percentage of error as well as coefficient determination. The obtained experimental results show that the neural model could predicate the tenacity with less than 3.5% error. Afterwards, utilizing genetic algorithms, a new method is proposed for optimal determination of input values in yarn production to reach the desired tenacity. We conducted several experiments for different ranges with various production cost functions. The proposed approach could find the best input values to reach the desired tenacity considering the production costs.

Web search engines are one of the most popular tools on the Internet which are widely-used by expert and novice users. Constructing an adequate query which represents the best specification of users’ information need to the search engine is an important concern of web users. Query expansion is a way to reduce this concern and increase user satisfaction. In this paper, a new method of query expansion is introduced. This method which is a combination of relevance feedback and latent semantic analysis, finds the relative terms to the topics of user original query based on relevant documents selected by the user in relevance feedback step. The method is evaluated and compared with the Rocchio relevance feedback. The results of this evaluation indicate the capability of the method to better representation of user’s information need and increasing significantly user satisfaction.

In this paper, we study WSN design, as a multi-objective optimization problem using GA technique. We study the effects of GA parameters including population size, selection and crossover method and mutation probability on the design. Choosing suitable parameters is a trade-off between different network criteria and characteristics. Type of deployment, effect of network size, radio communication radius, density of sensors in an application area, and location of base station are the WSN’s characteristics investigated here. The simulation results of this study indicate that the value of radio communication radius has direct effect on radio interference, cluster-overlapping, sensor node distribution uniformity, communication energy. The optimal value of radio communication radius is not dependent on network size and type of deployment but on the density of network deployment. Location of the base station affects radio communication energy, cluster-overlapping and average number of communication per cluster head. BS located outside the application domain is preferred over that located inside. In all the network situations, random deployment has better performance compared with grid deployment.