Dongshik Kang

Density Based Support Vector Machines for Classification

Support Vector Machines (SVM) is the most successful algorithm for classification problems. SVM learns the decision boundary from two classes (for Binary Classification) of training points. However, sometimes there are some less meaningful samples amongst training points, which are corrupted by noises or misplaced in wrong side, called outliers. These outliers are affecting on margin and classification performance, and machine should better to discard them. SVM as a popular and widely used classification algorithm is very sensitive to these outliers and lacks the ability to discard them. Many research results prove this sensitivity which is a weak point for SVM. Different approaches are proposed to reduce the effect of outliers but no method is suitable for all types of data sets. In this paper, the new method of Density Based SVM (DBSVM) is introduced. Population Density is the basic concept which is used in this method for both linear and non-linear SVM to detect outliers. Experiments on artificial data sets, real high-dimensional benchmark data sets of Liver disorder and Heart disease, and data sets of new and fatigued banknotes’ acoustic signals can prove the efficiency of this method on noisy data classification and the better generalization that it can provide compared to the standard SVM.

A new hierarchical clustering algorithm

The purpose of data clustering algorithm is to form clusters (groups) of data points such that there is high intra-cluster and low inter-cluster similarity. There are different types of clustering methods such as hierarchical, partitioning, grid and density based. Hierarchical clustering is a method of cluster analysis which seeks to build a hierarchy of clusters. A hierarchical clustering method can be thought of as a set of ordinary (flat) clustering methods organized in a tree structure. These methods construct the clusters by recursively partitioning the objects in either a top-down or bottom-up fashion. In this paper we present a new hierarchical clustering algorithm using Euclidean distance. To validate this method we have performed some experiments with low dimensional artificial datasets and high dimensional fMRI dataset. Finally the result of our method is compared to some of existing clustering methods.

Classification of Fatigue Bills Based on K-Means by Using Creases Feature

The bills in circulation generate a large amount of fatigue bills every year, causing various types of problems, such as the paper jam in automatic tellers due to overwork and exhaustion. A highly advanced bill classification technique, which distinguishes whether a bill is a reusable bill specifying the level of fatigue, is greatly required in order to comb out these problematic bills. Therefore, a purpose of this paper is to suggest a classification method of fatigue bills based on K-means with bill image data. The effectiveness of this approach is verified by the bill discriminant experimentation.

Comparative Study of Outlier Detection Algorithms for Machine Learning

Outliers are unusual data points which are inconsistent with other observations. Human error, mechanical faults, fraudulent behavior, instrument error, and changes in the environment are some reasons to arise outliers. Several types of outlier detection algorithms are developed and a number of surveys and overviews are performed to distinguish their advantages and disadvantages. Multivariate outlier detection algorithms are widely used among other types, therefore we concentrate on this type. In this work a comparison between effects of multivariate outlier detection algorithms on machine learning problems is performed. For this purpose, three multivariate outlier detection algorithms namely distance based, statistical based and clustering based are evaluated. Benchmark datasets of Heart disease, Breast cancer and Liver disorder are used for the experiments. To identify the effectiveness of mentioned algorithms, the above datasets are classified by Support Vector Machines (SVM) before and after outlier detection. Finally a comparative review is performed to distinguish the advantages and disadvantages of each algorithm and their respective effects on accuracy of SVM classifiers.

Prioritizing Road Maintenance Activities using GIS Platform and Vb.net

One of the most important factors for the sustainable development of any country is the quality and efficiency of its transportation system. The principled and accurate maintenance of roads, in addition to having a major impact on budget savings, improves the quality and service levels of the transportation system. For this reason, road management and maintenance are the main pillars of the transportation system in any country. Nowadays, due to the increased cost of maintaining roads and the lack of funding in this area, traditional ways of managing and maintaining roads, which are more based on the experience of the experts themselves, are no longer affordable. Hence, more recent, and more systematic methods have become more popular among relevant authorities. Afghanistan is a country facing problems such as budget deficits, lack of professional experts and advanced technology in road maintenance sector. This paper presents an example of using the GIS platform and vb.net to prioritize the road maintenance and rehabilitation activities based on identified criteria. A case study conducted in an academic environment and road maintenance and rehabilitation activities prioritized. The results show that the positive criterion has the greatest impact on the ranking of road maintenance activities. The characteristic of this process is to help the decision makers to plan road maintenance requirements to effectively and efficiently allocate funds for future planning.

Classification of Fatigue Bill Based on Support Vector Machine by Using Acoustic Signal

The bills in circulation generate a large amount of fatigue bills every year, causing various types of problems such as the paper jam in automatic tellers due to overworked and exhausted ones. An advanced technique is requested in order to classify the levels of fatigue as well as distinguish between the used and the new ones. Therefore, the purpose of this paper is to present the classification method of fatigue bills based on support vector machine(SVM) by using acoustic signals. The effectiveness of this approach is demonstrated by the bill identify experimentation based on the real acoustic signal.

A design of semantics filter using the inductive learning algorithm for GIS

We propose a semantics filter aimed at filtering unnecessary data for a user. In the Geographic Information Systems (GIS), the transfer of spatial data for each user is one of the important problems. We design a system that filters spatial data by using users information adoptively. Our system derives the degree of geographical knowledge by the fuzzy reasoning and decides the priority of buildings by the C4.5 using users information. Furthermore, in the process of the priority of buildings, we determine also the priority from geometrical and topological information. This paper describes a method for determining the priority of buildings.