Al-Rafidain University College For Sciences

the importance of study focuses on methods related to smoothing of Nonparametric Regression functions. This is for the purpose of producing the best methods convenient for various models. And for the Distribution Random error, in its Normal cases.
Thus, the most important purpose of the research, is to find what the studies so far, have offered in the field of Nonparametric Regression. Also to find alternative or modified methods; which are reliable for the treatment of conditions of failure regarding the methods in use, as well as to alleviate the some methods, especially those related to Bayesian procedures.
One of the most outstanding aims of the research focuses on the study of Nonparametric Regression using Bayesian variable selection. This suggests a modified technique to be reliable and of less complexity than the A simulation model has been performed for a number of models .To verify the performance of such methods, many criteria have been carried out.

Many researcher, authors and writers in the Iraqi ancient civilization face problems in writing documents which include cuneiform symbols. In many cases they leave blanks for the cuneiform symbols for an artists to draw the symbols on the final copy of the document, which is a tedious and erroneous process.
This paper introduce special word processor for above purpose, gives the ability of inclusion the cuneiform symbols, the fundamental operations of a word processor like file manipulation (load and save) ,ordinary language font size, shape and type selection, the copy paste operations , and print operations.
To keep simple, The user have the ability to load files produced by the introduced word processor in one of the powerful word processors like XP word, and use this word processor for a final and more sophisticated preparation of the document.

This paper presents a way of data combination of technique of several features and their mapping using discrete mltiwavelet transform (DMWT).This combination was tested for isolated-word speech recognition . It is shown that this approach introduce more accurate results . This is due to the use of MWT in the combination instead of putting several logic rules. This experiment considered as a good beginning in using multiwavelet in feature combination of speech signal. It was compared with method of linear combination applying to the same data which results in (87.75%) . For the DMWT gave (90.81%) . It is clear that the new method give much better performance than the conventional one.

Customer relationships management (CRM) is the most important business trend of the last decade , This new marketing philosophy call for a new relationship between the organization and its customers which aim at retaining this customers for attaining the long –term profit.
This analytical-theoretical study aim at heightening and framing the knowledge in this field , and to introduce an integrated model which enable the organization to use this concept and achieve its goals .

One of the most important factors of software engineering is to show how the testing strategies can be applied , which can be obtained by applying the testing strategies over the software model(s), the result of operation can be used as a guide to specify the performance and quality of application program . This system can test application program (such as Pascal , C, C++ programming language) by applying many methods and metrics testing strategies, these metrics are effort test metrics, stress test metrics, complexity test metrics , memory requirement test metrics , runtime test metrics , and coupling test metrics.

The idea of the mistakes becomes by form the exciting thinks are for the discussion whether it charging with the ascription is statisticians or the researchers have who many is the scopes and who trade with one another with this idea where believes all by that the subject of the mistakes it is from the difficult and complex thesis, and from here an idea came out this discussion and which deceives by pretenses come you explain the idea of the mistakes non random and simplified is by form and a statement of the differentiation between it and between the mistakes random to one from the most dangerous of the kinds of the mistakes non random made every effort and it is mistakes of the registration its appreciation cut and prediction in slices by its path and than my one from important kinds of the statistical analysis that is ((Markov Chain)) and the statement of an effect extent is the presence of the registration mistakes on the values of the Transition matrix elements and that is on the long run and a situation of the firmness and Stationary .In other words that the field of an interest is this discussion raise prediction highness amounting to the mistakes are non random in the matrix of the eventualities Transition the recalcitrant of the step and does this error fashion a danger to crash down when as its prediction accuracy my counting for the values of the Transition matrix and what is the form which will return to the matrix ? And then we study the estimation of the matrix , and it had supported the discussion with an application study about the products of the available milks in the markets.
The researcher is according to an information then that this study becomes the first on both local and world levels.

Wavelet Transform in 1992 recently as a combination with elman recurrent neural network. These are found very useful for time scale representation and widely used in signal processing and computer vision. In this research proposed an algorithm to reconstruct image that had lost some of its parts or uncompleted and some of wavelet transform properties in analysis to represent things. The algorithm depend on 8-level of wavelet decomposition with feature extraction and elman recurrent neural network for uncompleted image in small and big losing blocks. The reconstruction done in 100% accurate with the existing of original image even if the losing is 75% from the original image.

This paper focuses on implementing genetic algorithm on distributed memory. This parallel computer can be programmed using MPI standard message passing interface. We use a DAG to model parallel computation. The work is based on priority task graph representation of jobs that contain sequential segments with varying dependencies. We consider compile-time static scheduling when communication overhead is not negligible. DSC is used to cluster tasks and then use a load balancing and physical mapping heuristic to map the clusters onto processors. The main optimization issues are balancing computation among processors, reducing inter-processor communication and overlapping communication with computation. Theoretical and experimental results are presented to verify the performance of these algorithms.

Advance layout problem (Alp) is used to search for an optimal layout of machines. This Research Describes a novel method, based on genetic algorithms (GA) to solve the machine layout problem, where developing machine layout is an important step in designing manufacturing , renovation of factories, distribution centers, hospitals, banks, department stores, military supply, depots, university, ect .The research studies the problems of adding the heterogeneous objects, continuous placement in the general spatial layout problem. Results are achieved through the use of local optimizer, separation algorithm with genetic algorithm also called hybrid simple genetic. Results show the potentiality of the proposed algorithm in solving the problem and outperforming previous algorithms.

The World Wide Web (WWW) allows people to share information globally. The amount of information grows without bound. In order to extract information that we are interested in, we need a tool to search the Web. The tool is called a search engine. This research aims to build a ranking system to retrieve the most relevant Web pages with the queries that are presented by the users; the ranking system depends on Web pages content and their hyperlinks’ structure. To retrieve the optimal results, the modified ranking system consists of two main subsystems. The first subsystem is Term-based Ranking subsystem that assigns a rank score to the Web pages depending on their contents by implementing TF-IDF Algorithm and Vector Spreading Activation Algorithm. And the second subsystem is Link-based Ranking subsystem that assigns a rank score to the Web pages depending on their hyperlinks by implementing an improved PageRank algorithm. Enhanced Indexing Process is considered as a main tool that supports in retrieving best results.

A general new technique is used to evaluate the determinant value for the matrix of order NXN by using a reduction determinant value of any order (t) and then use it to solve the system of N linear equations by using Grammar's rule (the matrix of the system of order N*N must be square matrix) . Flowchart has been designed to explain the technique.

The program was written by using visual basic language to apply our new approach, some examples are solved as an implementation of the new technique program; also the program can be setup on any computer as a ready program for determinant & use it to solve the system of linear equations.

Transport calculations of photon and neutron fluxes and doses were made. The Monte Carlo multigroup experimental transport code MORSE-CG [1] was used for these calculations with its modified version [2].
Distributions of gamma ray fluxes through the shield of a pressurized water reactor were obtained. The shield was represented by layers of water, steel and lead; it was a semi realistic representation of a working reactor.
A simulation was made for the total dose that could be received by a person living near a very active radiating source of neutrons or photons.
The obtained results were compared with experimental referenced data which shows a very good agreement.

The main purpose of this paper is divided into two parts: First is the classification of the partial integro-differential equation(PIDE).Second is the finding the approximate solution for this equation by using Least- square and Collocation methods.

Data mining, which is defined as the process of extracting previously unknown knowledge, and detecting intersecting patterns from a massive set of data, has been a very active research. Image mining is more than just an extension of data mining to image domain. It is an interdisciplinary endeavor that draws upon expertise in computer vision, image processing, image retrieval, data mining, machine learning, database, and artificial intelligence. Despite the development of many applications and algorithms in the individual research fields cited above, research in image mining is still in its infancy.

In this research we proposed a method to find association rules between the objects in images using image mining. The idea of the proposed method is selecting a collection of images that belong to a specific field, after the selection stage we will extract the objects from each image and indexing all the images with its objects in transaction database, the data base contain image identification and the objects that belong to each image with its features.
After creating the transaction data base that contains all images and their features we will use the data mining method to associate rules between the objects. This will help us for prediction. The steps of the work include extract objects from images using object extraction and analysis which is a field of image processing. After extracting the object, its value will be normalized and create a suitable data base for image mining to find the frequent item sets between these objects to associate rules between them.

Transport calculations were made for radiation doses and fluxes in different materials for different ranges. The Monte Carlo multigroup experimental transport code MORSE-CG [1] was used for these calculations. Some modifications were introduced to this code. These modifications include the addition of subroutine called SOURCE, and the addition of a function called DIREC. The subroutine SOURCE allows the radiation source of particles to have certain geometry instead of the built-in point source. While the function DIREC makes the radiation to transport in a certain direction to reduce the time of calculations.
The obtained results of calculations were compared with experimental referenced data which show a good agreement.

Knowledge-based economy becomes a common place among policy markers and main stream economists. The shift to a digital knowledge-based economy, prompted by new goods and services, will be a powerful engine for growth, competitiveness and jobs. In addition, it will improve citizen’s quality of life and environment. It enables us to discard some meanings and practices and retain others. Knowledge-based economy exhibits a dynamics different from a market economy. No organization can remain static and survive. Internal and external forces will provide or destroy opportunity and only those organizations equipped with technological infrastructure and strategy to economically exploit all facets of their knowledge will flourish

Genetic algorithms are very efficient at exploring the entire search space; however, they are relatively poor at finding the precise local optimal solution in the region at which the algorithm converges. Hybrid genetic algorithms are the combination of learning algorithm (Backpropagation), usually working as evaluation functions, and genetic algorithms. There are two basic strategies in using hybrid GAs, Lamarckian and Baldwinian evolution. Traditional schema theory does not support Lamarckian learning, i.e., forcing the genetic representation to match the solution found by the learning algorithm. However, Lamarckian learning does alleviate the problem of multiple genotypes mapping to the same phenotype. Baldwinian learning uses learning algorithm to change the fitness landscape, but the solution that is found is not encoded back into the genetic string. We presented hybrid genetic algorithms for optimizing weights as well as the topology of artificial neural networks, by introducing the concepts of Lamarckian and Baldwin evolution effects. Experimental results with extensive set of experiments show that the hybrid genetic algorithm exploiting the Baldwin effect more effect than Lamarckian evolution but is slow in convergence, and The results of the proposed algorithms outperformed those of the previous algorithms.