IRANIAN COMMUNICATION AND INFORMATION TECHNOLOGY
Year:2016 | Volume:7 | Issue:25-26|Papers Count:6

One of the machine learning tasks is supervised learning. In supervised learning we infer a function from labeled training data. The goal of supervised learning algorithms is learning a good hypothesis that minimizes the sum of the errors. A wide range of supervised algorithms is available such as decision tress, SVM, and KNN methods. In this paper we focus on decision tree algorithms. When we use the decision tree algorithms, the data is partitioned by axis- aligned hyper planes. The geometric concept of decision tree algorithms is relative to separability problems in computational geometry. One of the famous problems in separability concept is computing the maximum bichromatic discrepancy problem. There exists an -time algorithm to compute the maximum bichromatic discrepancy in d dimensions. This problem is closely relative to decision trees in machine learning. We implement this problem in 1, 2, 3 and d dimension. Also, we implement the C4.5 algorithm. The experiments showed that results of this algorithm and C4.5 algorithm are comparable.

By the increasing of video communication in portable and functional devices, encoders design with low complexity and high performance are required. H.264 / AVC standard offers higher compression efficiency than previous standards. But this standard by employing several powerful coding techniques, considerably increased complexity at the encoder. This paper presents a new algorithm to reduce the complexity of the H.264/AVC encoder. The proposed method uses simple directional masks, neighboring blocks modes and detection of 4´4 and/or 16´16 intra estimation modes with determination of quantization parameters for fast mode selection in Intra-Frame Modes prediction. Experimental results show that the proposed method reduces maximum 29% of the encoding time, while has little effect on visual quality and PSNR.

Reduction of the quality of the image formed by an optical system is a function of different parameters such as lens aberrations, CCD digitization errors, and the errors of system assembling. Assembling errors usually consist of two types: 1) the prism error, which is the error of non-orthogonality of the optical axis and the image plane 2) the de-centering error, which is error of not passing the lens optical axis through the center of the image plane. This paper attempts to correct the blind of the lens aberration and the de-centering error. To this end, Seidel aberrations are rewritten in the form of the modified Zernike moments based on the second kind Chebyshev polynomials as discrete functions on the Cartesian space. Then, the modified moments reformulated to model de-centered phase aberration function by considering the de-centering error. Finally, the model parameters are divided into two classes of symmetric and asymmetric ones. Then, these parameters are estimated through poly-spectral analysis, i.e., bi-coherence and tri-coherence analysis, respectively. Experimental results confirm the accuracy and efficiency of the proposed solution.

Learning to rank refers to machine learning techniques for training a model in a ranking task. Learning to rank has been shown to be useful in many applications of information retrieval, natural language processing, and data mining.Learning to rank can be described by two systems: a learning system and a ranking system. The learning system takes training data as input and constructs a ranking model. The ranking system then makes use of the learned ranking model for ranking prediction. In this paper, a new learning algorithm based on ensemble learning for learning ranking models in information retrieval is proposed. This algorithm iteratively constructs weak learners using a fraction of the training data whose weight distribution is determined based on previous weak learners. The proposed algorithm combines the weak rankers to achieve the final ranking model. This algorithm constructs a ranking model on a fraction of the training data to increase the accuracy and reduce the learning time. Experimental results based on Letor.3 benchmark dataset shows that the proposed algorithm significantly outperforms other ensemble learning algorithms.

The required factors have been investigated to apply knowledge management by carrying out the field study using staff opinions of Kermanshah Telecommunication Company in this research. After indicating the required factors of applying knowledge management, relationship between each factor of knowledge process system, organizational culture and information technology system with the knowledge management have been analyzed statistically by the obtained results using SPSS program.Kermanshah Telecommunication Company has a population of 370 staff that a sample of 77 staff has been used. It has been found that there are acceptable correlation between knowledge process system, organizational culture and information technology system with the knowledge management which correlation coefficient (R) of each relationship is 0.854, 0.915 and 0.812 respectively. There are some different between the effects of knowledge process system, organizational culture and information technology system on the applying knowledge management as the effects knowledge process system, information technology system and organizational culture are in the fist, second and third step respectively.