Data Mining

Assessment of future precipitation variations prevailing in an area is essential for the research regarding climate and climate change. The current paper focuses on 3 selected areas in Greece that present different climatic characteristics due to their location and aims to assess and compare the future variation of annual and seasonal precipitation. Future precipitation data from the ENSEMBLES anthropogenic climate-change (ACC) global simulations and the Climate Local Model (CLM) were obtained and analyzed. The climate simulations were performed for the future periods 2021-2050 and 2071-2100 under the A1B and B1 scenarios. Mann-Kendall test was applied to investigate possible trends. Spatial distribution of precipitation was performed using a combination of dynamic and statistical downscaling technique and Kriging method within ArcGIS 10.2.1.
The results indicated that for both scenarios, reference periods and study areas, precipitation is expected to be critically decreased. Additionally, Mann-Kendall test application showed a strong downward trend for every study area. Furthermore, the decrease in precipitation for the Ardas River basin characterised by the continental climate will be tempered, while in the Sperchios River basin it will be smoother due to the influence of some minor climatic variations in the basins' springs in the highlands where milder conditions occur. Precipitation decrease in the Geropotamos River basin which is characterized by Mediterranean climate will be more vigorous. B1 scenario is appeared more optimistic for the Ardas and Sperchios River basins, while in the Geropotamos River basin, both applied scenarios brought similar results, in terms of future precipitation response.

The complexity of the medical diagnosis is faced by practitioners relying mainly on their experiences. This can be acquired during daily practices and on-the-job training. Given the complexity and extensiveness of the subject, supporting tools that include knowledge extracted by highly specialized practitioners can be valuable. In the present work, a Decision Support System (DSS) for hand dermatology was developed based on data coming from a Visit Report Form (VRF). Using a Bayesian approach and factors significance difference over the population average for the case, we demonstrated the potentiality of creating an enhanced VRF that include a diagnoses distribution probability based on the DSS rules applied for the specific patient situation.

In this paper we present a method for automatic detection of visual patterns in a given news video format by investigating similarities in a set of videos of that format. The approach aims at reducing the manual effort needed to create models of news broadcast formats for automatic video indexing and retrieval. Our algorithm has only very few parameters and

The web is an enormous information space where a large number of an individual article or unit such as documents,
images, videos or other multimedia can be retrieve. In this context, several information technologies have been
developed to assist users to gratify their searching needs on web, and the most used by users are search engines as
Yahoo, Google, Netscape, e-Bay, e-Trade, Expedia, Amazon, Bing, Ask, and so on. The search engines allow users to find
web relevant resources by setting up their queries and reviewing a list of answers. In this paper a search result optimization
method for search engine optimization by page rank updating, query recommendation and query reformulation are
proposed. It explores the users queries registered in the search engine's query logs in order to learn how users search and
also in order to design algorithms that can improve the correctness of the answers suggested to users. The proposed
method starts by exploring the query logs to find query clusters and identify session of queries then it examine query logs
to discover useful relationship among pages, keywords and queries within clusters using association rule mining
algorithms such as an apriori algorithm and automated apriori algorithm. The authors also showed that automated
apriori algorithm generates more strong rules as compare to apriori algorithm.
Keywords- Web Mining, Apriori Algorithm, Automated Apriori Algorithm, Clustering, Rank Improvement Algorithm, Page,
Keyword, and Query Association

One of the important discussions in data mining is extracting effective and useful rules from the great set of datasets. So, we should follow set of features that at first; are without any noise; secondly, having a little correlation with other features. In other words, we should use instances that are distinctive with other features. So, in this paper we present a combined approach to consider how factors such as distinct features and instances are useful for extracting the rules. In this approach we used a trained neural network to explore useful features, clustering to find out the best instances from dataset and finally we used artificial immune system for rules extraction. In order to evaluating of our introduced approach, we applied it on the UCI dataset of breast cancer diagnosis. Our experiments demonstrate that the combined proposed approach generates reliable rules and contributes more accuracy eventually; these results show the proposed method has %5.9 better accuracy relative to CART method.

In the most of standard learning algorithms it is presumed or at least expected that distributions governing on different classes of at-hand dataset are balanced; it means that there are the identical number of data points in each class. It is also resumed there that the misclassification cost of each data point is a fixed value regardless of its class. The standard algorithms fail to learn at
the imbalanced datasets. An imbalance dataset is the one that the distributions governing among their classes are not identical for all classes. A very well-known domain example of imbalanced datasets is automatic patient detection. In such systems there are many clients while a few of them
are patient and the others are all healthy. So it is very common to face an imbalanced dataset in a system for patient detection. In a breast cancer patient detection that is a special case of the
mentioned systems, we try to discriminate the patient clients from healthy clients. It should be noted that the imbalanced shape of a dataset can be relative where the mean number of samples is high in the minority class, but it is very less than the number of samples in the majority class. This paper presents an algorithm which is well-suited to the field of non-relative imbalanced datasets, in both speed and efficacy of learning. The experimental results show that the performance of the
proposed algorithm outperforms some of the best methods in the literature.

We develop a linear time method for transforming clusters of 2D-point data into area data while identifying the shape robustly. This method translates a data layer into a space filling layer where shaped clusters are identified as the resulting regions. The method is based on robustly identifying cluster boundaries in point data using the Delaunay Diagram. The method can then

... Gaurav N. Pradhan Dept. of Computer Science University of Texas at Dallas Richardson, TX 75080. gaurav@utdallas.edu Balakrishnan Prabhakaran Dept. of Computer Science University of Texas at Dallas Richardson, TX 75080. praba@utdallas.edu ...