• The KIPS Transactions:PartB
• /
• v.10B no.6
• /
• pp.719-724
• /
• 2003

This paper proposed a heuristic to enhance the speed of kNN without loss of its accuracy. The proposed heuristic minimizes the computation of the similarity between two documents which is the dominant factor in kNN. To do this, the paper proposes a method to calculate the upper limit of the similarity and to sort the training documents. The proposed heuristic was implemented on the existing framework of the text categorization, so called, AI :: Categorizer and it was compared with the conventional kNN with the well-known data, Router-21578. The comparisons show that the proposed heuristic outperforms kNN about 30∼40% with respect to the execution time.

• Journal of Internet Computing and Services
• /
• v.5 no.5
• /
• pp.129-138
• /
• 2004

In the telecommunications network, multicasting is widely used recently. Multicast tree problem is modeled as the NP-complete Steiner problem in the networks. In this paper, we study algorithms for finding efficient multicast trees with hop and node degree constraints. Multimedia service is an application of multicasting and it is required to transfer a large volume of multimedia data with QoS(Quality of Service). Though heuristics for solving the multicast tree problems with one constraint have been studied. however, there is no optimum algorithm that finds an optimum multicast tree with hop and node degree constraints up to now. In this paper, an approach for finding an efficient multicast tree that satisfies hop and node degree constraints is presented and the experimental results explain how the hop and node degree constraints affect to the total cost of a multicast tree.

• Journal of Computing Science and Engineering
• /
• v.1 no.2
• /
• pp.162-176
• /
• 2007

Machine Learning (ML) has become very popular within Data Mining (KDD) and Artificial Intelligence (AI) research and their applications. In the ML and KDD contexts, two main approaches can be used for inducing a Bayesian Network (BN) from data, namely, Conditional Independence (CI) and the Heuristic Search (HS). When a BN is induced for classification purposes (Bayesian Classifier - BC), it is possible to impose some specific constraints aiming at increasing the computational efficiency. In this paper a new CI based approach to induce BCs from data is proposed and two algorithms are presented. Such approach is based on the Markov Blanket concept in order to impose some constraints and optimize the traditional PC learning algorithm. Experiments performed with the ALARM, as well as other six UCI and three artificial domains revealed that the proposed approach tends to execute fewer comparison tests than the traditional PC. The experiments also show that the proposed algorithms produce competitive classification rates when compared with both, PC and Naive Bayes.

• IEIE Transactions on Smart Processing and Computing
• /
• v.5 no.6
• /
• pp.430-439
• /
• 2016

Estimation and analysis of traffic jams plays a vital role in an intelligent transportation system and advances safety in the transportation system as well as mobility and optimization of environmental impact. For these reasons, many researchers currently mainly focus on the brilliant machine learning-based prediction approaches for traffic prediction systems. This paper primarily addresses the analysis and comparison of prediction accuracy between two machine learning algorithms: Naïve Bayes and K-Nearest Neighbor (K-NN). Based on the fact that optimized estimation accuracy of these methods mainly depends on a large amount of recounted data and that they require much time to compute the same function heuristically for each action, we propose an approach that applies multi-threading to these heuristic methods. It is obvious that the greater the amount of historical data, the more processing time is necessary. For a real-time system, operational response time is vital, and the proposed system also focuses on the time complexity cost as well as computational complexity. It is experimentally confirmed that K-NN does much better than Naïve Bayes, not only in prediction accuracy but also in processing time. Multi-threading-based K-NN could compute four times faster than classical K-NN, whereas multi-threading-based Naïve Bayes could process only twice as fast as classical Bayes.

• The Journal of Information Technology
• /
• v.3 no.2
• /
• pp.73-85
• /
• 2000

The efficient algorithms are suggested in this study for solving the multicommodity network flow problems applied to Communications Systems. These problems are typical NP-complete optimization problems that require integer solution and in which the computational complexity increases numerically in appropriate with the problem size. Although the suggested algorithms are not absolutely optimal, they are developed for computationally efficient and produce near-optimal and primal integral solutions. We supplement the traditional Lagrangian method with a price-directive decomposition. It proceeded as follows. First, A primal heuristic from which good initial feasible solutions can be obtained is developed. Second, the dual is initialized using marginal values from the primal heuristic. Generally, the Lagrangian optimization is conducted from a naive dual solution which is set as ${\lambda}=0$. The dual optimization converged very slowly because these values have sort of gaps from the optimum. Better dual solutions improve the primal solution, and better primal bounds improve the step size used by the dual optimization. Third, a limitation that the Lagrangian decomposition approach has Is dealt with. Because this method is dual based, the solution need not converge to the optimal solution in the multicommodity network problem. So as to adjust relaxed solution to a feasible one, we made efficient re-allocation heuristic. In addition, the computational performances of various versions of the developed algorithms are compared and evaluated. First, commercial LP software, LINGO 4.0 extended version for LINDO system is utilized for the purpose of implementation that is robust and efficient. Tested problem sets are generated randomly Numerical results on randomly generated examples demonstrate that our algorithm is near-optimal (< 2% from the optimum) and has a quite computational efficiency.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.7
• /
• pp.87-95
• /
• 2012

In this paper, PTAS three-dimensional Steiner minimum tree connecting numerous input nodes rapidly in 3D space is proposed. Steiner minimum tree problem belongs to NP problem domain, and when properly devised heuristic introduces, it is generally superior to other algorithms as minimum spanning tree affiliated with P problem domain. But when the number of input nodes is very large, the problem requires excessive execution time. In this paper, a method using PTAS is proposed to solve the difficulty. In experiments for 70,000 input nodes in 3D space, the tree produced by the proposed 8 space partitioned PTAS method reduced 86.88% execution time, compared with the tree by naive 3D steiner minimum tree method, though increased 0.81% tree length. This affirms the proposed method can work well for applications that many nodes of three dimensions are need to connect swifty, enduring slight increase of tree length.