روش طبقه بندی جدید متاهیوریستیک الهام گرفته از خفاش برای داده های میکرو آرایه
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی|
|8064||2012||5 صفحه PDF||8 صفحه WORD|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Procedia Technology, Volume 4, 2012, Pages 802–806
2. کار مربوطه
3. نمایش شماتیک مدل پیشنهادی
4. شیوهی کار BAT
شکل 2. خفاش سیگنال صوتی با فرکانس f ارسال میکند.
شکل 3. سیگنال اکو برای محاسبهی فاصله S استفاده میشود.
4.1 محاسبه فرکانس
4.2 محاسبه فاصله
4.3 به روز رسانی موقعیت خفاش
4.4 به روز رسانی فرکانس f و وزن پس از تغییر موقعیت خفاش
5. ارزیابی تجربی و نتیجه
6. نتیجه گیری و کارهای آینده
The main objective of a classifier is to discover the hidden class level of the unknown data. It is observed that data size, number of classes and dimension of feature space and inter class separability affect the performance of any classifier. For a long time, efforts are made in improving efficiency, accuracy and reliability of classifiers for a wide range of applications. Different optimization algorithms such as Particle Swarm Optimization (PSO) and Simulated Annealing (SA) have been used to enhance the accuracy of classifiers. Bat is also a metaheuristic search algorithm which is use to solve multi objective engineering problem. In this paper, a model has been proposed for classification using bat algorithm to update the weights of a Functional Link Artificial Neural Network (FLANN) classifier. Bat algorithm is based on the echolocation behaviour of bats. The proposed model has been compared with FLANN, PSO-FLANN. Simulation shows that the proposed classification technique is superior and faster than FLANN and PSO-FLANN.
High dimensionality of microarray data sets is a crucial issue to be considered while designing classifiers . To handle the curse of high dimensionality, the data sets need to be pre-processed by reducing the redundant and irrelevant features. By removing such features or attribute we can also reduce the computational complexity. Principal Component Analysis (PCA) is used to deal with curse of dimensionality for micro array data set. The ultimate goal of any pattern recognition system is to achieve the best possible classification performance for a given problem domain. Meta heuristic algorithms like PSO  and SA are the powerful methods for solving many optimization problems. The fine adjustment of the parameters of the above techniques enhances the accuracy of the classifiers. In this paper, bat algorithm is used to update the weights of a FLANN classifier. Bat emits sound of various wavelength and frequency in the search of prey and direction . Bat flies with velocity v at position x with different sound frequency f. Bat adjusts its velocity, direction and frequency on hearing echo signal. In this paper, a new meta heuristic bat algorithm has been formulated and also the whole working principle of the algorithm is explained. This paper is organized as follows; section 2 describes the related work, section 3 shows the schematic representation of the proposed model, section 4 contains the workingprocedure of bat, section 5 gives the experimental evaluation and result; finally, section 6 deals with conclusion and future work.
نتیجه گیری انگلیسی
In this paper, the bat algorithm successfully formulated and is used to update the weight of the FLANN classifier. Wide knowledge of bats echolocation signals and their specific features results in a good accuracy in FLANN. From the formulation of the bat algorithm, its implementation and comparison it has been observed that it is a very promising algorithm. It is more powerful than PSO. The primary reason is bat algorithm uses a good combination of major advantages of PSO. In future, it can be used for classifier fusion. A natural extension to the current bat algorithm can be used in other engineering application areas.