هذا النظام يا اخي يقوم بعمل ما يسمى في النهايه توقع من خلال الرجوع الى قاعدة البيانات يوهذا يعني بعد ان نقوم بتدريب البيانات من خلال ما يسمى normalaization وبعدها باقي الخطوات اولا اخي ارجوا منك الاطلاع على هذا الشرح الموجود في المرفقات والذي يوجد به ايضا المستويات المخفيه التي تحدثت عنها وموجوده بالرسم والشرح ايظا ثم بعد ذالك سوف احاول ان اساعدك على فهم ما تريد في المرات القادمه بعد ان تقرا ما هو موجود وتقبل مني كل الاحترام والتقدير ..........
اولا افهم الموضوع ثم ابدا بالبرمجه في ما بعد
GLASS CLASSIFICATION SYSTEM
In a variety of fields, researchers (inspired by bioinformatics research area) are applying neural network models to solve problems that so far have not been solvable using early architectures or models. The aim of the project is to perform an improvement study of Neural Network models available for classification, a Multi Layer Perceptron feed forward trained by Back propagation algorithm to solve the classifying glass data problem.
INTRODUCTION المقدمه
Among the many techniques and models are utilized in data mining one of them is classification. Classification is a supervised process, that consist of the problem is to label the records which are as yet unlabelled. The labeled or "training" records are used to learn the attributes of a group, which in turn are used to label the new record. Which the categories in classification are set externally.
NEURAL NETWORK
Artificial neural networks (ANNs) simulate the concepts developed through physiological modeling of the human brain in computational mechanics. ANNs contain a set of highly interconnected processing elements that are constructed in a regular architecture and act in parallel. The overall behavior of an ANN exhibits the abilities of learning, recalling, and generalization from training patterns or data by adjusting theconnection weights inside the network. The model of an ANN includes three basic entities: functions of the processing elements, the network topology, and learning rules (methods used to store information in the network). For a multilayer feed forward network topology, the error back-propagation learning algorithm is one of the most widely used supervised learning methods. The algorithm can perform rather high-quality generalization by escaping from local minima provided that the number of hidden nodes is properly selected. However, because the error surface may contain lots of areas with shallow slopes in multiple dimensions, the algorithm tends to converge rather slowly.
Neural networks have emerged as an important tool for classification. The recent vast research activities in neural classification have established that neural networks are a promising alternative to various conventional classification methods. The advan***e of neural networks lies in the following theoretical aspects. First, neural networks are data driven self-adaptive methods in that they can adjust themselves to the data without any explicit specification of functional or distributional form for the underlying model. Second, they are universal functional approximates in that neural networks can approximate any function with arbitrary accuracy (G. Cybenko, 1989) ,( K. Hornik, 1991),
Thus ANNis an information processing system that has certain performance characteristics in common with biological neural networks. ANNhas been developed as generalizations of mathematical models of human cognition or neural biology, ****d on the assumptions that:1. Information processing occurs at simple elements called neurons.
2. Signals are passed between neurons over connection links.
3. Each connection link has an associated weight, which in a typical neural Net multiplies the signal transmitted.
4. Each neuron applies an activation function (usually non-linear) to its net
Input (sum of weighted input signals) to determine its output signal.
A neural network is also characterized by the following characteristics: 1. Its pattern of connections between the neurons (called its architecture). 2. Its method of determining the weights on the connections (called its Training, or learning, algorithm). 3. Its activation function.
The ANN architecture consists of a large number of simple processing elements called neurons, units, cells or nodes. Each neuron is connected to other neurons in links which have a weight associated with them. The weights represent the information being used by the network to solve the problem. Thus ****d on the problem (classifying Iris plants), attempting to solve the interconnections and the weights associated vary. The initial assignment of weights is a topic of important discussion, as the closer the initial weights are to the final weights the faster the network is trained and ready to use.
In general neural network is composed of groups of functionally associated neurons a single neuron can be connected to many other neurons and the total number of neurons and connections in a network can be very large. There are tow type from network layer, the first one a single layer perception network consists of one or more artificial neurons in parallel.
The demonstration of the limitations of single layer neural network was a single significant factor in the decline in interest in neural networks in the 1970's. The discovery (by several researchers independently) and widespread dissemination f an effective general method of training a multi layer neural network (Rumelhart, Hinton & Williams, 1986). The very general nature of the back propagation training method means that a back propagation net (a multi layer, feed forward net trained by back propagation) can be used to solve problems in many areas. The training of a network by back propagation involves three s***es: the feed forward of the input training pattern, the calculation and back propagation of the error and the adjustment of weights. Although a single layer net can learn it is severely limited in its mapping. A multi layered net can be mapped to solve any problem up to any arbitrary accuracy.
Knowledge acquisition by extraction of logical rules from the sample data is an important and difficult problem in computational intelligence. Neural networks, in particular multi-layered perceptrons (MLPs), are useful classifiers that can learn arbitrary vector mappings from the input to the output space and successfully use this mapping in novel situations. The knowledge acquired by neural systems is represented in a set of numerical parameters and architectures of these networks in a way that is incomprehensible for humans. Some classification problems have an inherent logical structure and even in other cases it may be preferable to use logical rules instead of adaptive classifiers. Many methods to analyze trained neural networks, extract logical rules and select classification features have been devised in the past. These methods focus on analysis of parameters (weights) of trained networks, trying to achieve high fidelity of performance, i.e. obtaining identical classification results by extracted logical rules in comparison to the original networks. Analysis of complex networks is quite difficult and may lead to a large number of rules, too large to be useful in practice. Non-standard form of rules, such as M of N (M out of N antecedents should be true) or decision trees, are sometimes useful. There are tow type from machine learning in NN, supervised learning machine learning technique for creating a function from training data. The training data consist of pairs of input ******s and desired outputs. The output of the function can be a continuous value or can predict a class label of the input ****** called classification. The task of the supervised learner is to predict the value of the function for any valid input ****** after having seen a number of training examples, to achieve this; the learner has to generalize from the presented data to unseen situations in a reasonable way. The unsupervised learning machine learning where a model is fit to observations. It is distinguished from supervised learning by the fact that there is no a priori output. In unsupervised learning, a data set of input ******s is gathered. Unsupervised learning then typically treats input ******s as a set of random variables. A joint density model is then built for the data set. GLASS CLASSIFICATION SYSTEM عمل النظام
Here we will develop system to analysis the elements which enter In glass structure then inference the glass type if it is building glass or industry glass the study of classification of types of glass by analysis these chemical elements, so we will collect the data of chemicals element which enter in glass characterize, then analysis this data then relate every percents with its produced glass. The glass elements في المرفقات
1- (Id) number: 1 to 214
2 - (RI) refractive index
3 - (Na) Sodium
4 - (Mg) Magnesium
5 - (Al)Aluminum
6 - (Si) Silicon
7 - (K) Potassium
8 - (Ca) Calcium
9 - (Ba) Barium
10 - (Fe) Iron
The Number of Instances is 214 and the Number of Attributes is 10 And we identify the type of glass:
1- Building glass
2-industry glass
في ما بعد يمثل كيف نقوم بصياغة المستويات الخفيه والتي رمزت لها ب HI,H2 ومستوى الادخال الذي يتكون من جميع المكونات التي تدخل في صناعة الزجاج الصناعي او الكونات الطبيعيه في الزجاج الطبيعي , اما الخرجات فتمثل نوعيه الزجاج المتوقه بع ان نقوم بادخال من ضمن البرنامج وتكون المواد بنسب محدد ويقوم النظام بادخال هذه المكونات وعمل كل الخوارزميات الموجوده فيه ومن ثم يخرج نوع الزجاج المتوقع بناء على المدخلات ومقارنتها في ما قام به النظام من تدرب عليه النظام من خلال قاعدة البيانات الموجوده اصلا عن نفس المكونات سوف تجده على شكل رسم مع المرفقات .
In this system we have many forms that do the experiments and help the user to know and solve the glass structure then inference the glass type if it is building glass or industry glass, however, the first form or the main form include buttons.
In the final s***e, make the past process in these values the system will provide us the goal from this values, that will make us make truth decisions, reduce the cost and lose time, we can know how. When and why will us product the glass.
CONCLUSION ما ذا يعني الذي قمنا به
By use ANN we develop system to solve the problem which front use in glass classification related with the chemistry elements that be in its formed the glass, by use ANN we can reduce the search and comparing between elements to arrive to the fit way to fined fit product, we can arrive to correct decision in easy and fast way.
اشهد ان لا اله الا الله واشهد ان محمد رسول الله
انا اسف لرداءة الكتابه ولكن عذرا لاني اكتب مباشرة على مربع الردود اي بمعنا اخر لا اقوم بعمل مسوده مبدئيه ولكن للاسف كل هذا لضيق الوقت ولكن مع ذالك حاولت ان اقدم لك ما اقدر عليه الان حتى ولو كان هذا الامر على حساب الوقت لدي 
.
الله المستعان ارفق صورة
للتسجيل اضغط
هـنـا
10-13-2011
غير مجاب: طلب مساعدة من طلاب علوم حاسبات
المواضيع المتشابهه
1اعجابات
العرض العادي
