Data Mining Procedures Research Articles

ONCO-i2b2: improve patients selection through CBR techniques with heterogeneous distance functions

ONCO-i2b2: improve patients selection through CBR techniques with heterogeneous distance functions

Integrating Constitutive Gene Expression and Chemoactivity: Mining the NCI60 Anticancer Screen

Studies into the genetic origins of tumor cell chemoactivity pose significant challenges to bioinformatic mining efforts. Connections between measures of gene expression and chemoactivity have the potential to identify clinical biomarkers of compound response, cellular pathways important to efficacy and potential toxicities; all vital to anticancer drug development. An investigation has been conducted that jointly explores tumor-cell constitutive NCI60 gene expression profiles and small-molecule NCI60 growth inhibition chemoactivity profiles, viewed from novel applications of self-organizing maps (SOMs) and pathway-centric analyses of gene expressions, to identify subsets of over- and under-expressed pathway genes that discriminate chemo-sensitive and chemo-insensitive tumor cell types. Linear Discriminant Analysis (LDA) is used to quantify the accuracy of discriminating genes to predict tumor cell chemoactivity. LDA results find 15% higher prediction accuracies, using ∼30% fewer genes, for pathway-derived discriminating genes when compared to genes derived using conventional gene expression-chemoactivity correlations. The proposed pathway-centric data mining procedure was used to derive discriminating genes for ten well-known compounds. Discriminating genes were further evaluated using gene set enrichment analysis (GSEA) to reveal a cellular genetic landscape, comprised of small numbers of key over and under expressed on- and off-target pathway genes, as important for a compound’s tumor cell chemoactivity. Literature-based validations are provided as support for chemo-important pathways derived from this procedure. Qualitatively similar results are found when using gene expression measurements derived from different microarray platforms. The data used in this analysis is available at http://pubchem.ncbi.nlm.nih.gov/and http://www.ncbi.nlm.nih.gov/projects/geo (GPL96, GSE32474).

Improving Classifications of Medical Data Based on Fuzzy ART2 Decision Trees

Analyzing given medical databases provide valuable references for classifying other patients symptoms. This study presents a strategy for discovering fuzzy decision trees from medical databases, in particular Harbeman's Survival database and the Blood Transfusion Service Center database. Harbeman's Survival database helps doctors treat and diagnose a group of patients who show similar past medical symptoms and the Blood Transfusion Service Center database advises individuals about when to donate blood. The proposed data mining procedure involves neural network based clustering using Adaptive Resonance Theory 2 (ART2), and the extraction of fuzzy decision trees for each homogeneous cluster of data records using fuzzy set theory. Besides, another objective of this paper is to examine the effect of the number of membership functions on building decision trees. Experiments confirm that the number of erroneously clustered patterns is significantly reduced compared to other methods without preprocessing data using ART2.

Some properties of evaluated implications used in knowledge-based systems and data-mining

The core of expert knowledge is typically represented by a set of rules (implications) assigned with weights specifying their (un)certainties. The task of inference mechanism in such rule- based expert systems can be analyzed from the many-valued (fuzzy) logic perspective. On the other hand, implicational relations between two Boolean attributes derived from data (association rules) are quantified in data-mining procedures by (0,1)-valued functions defined on four-fold tables corresponding to pairs of the attributes. In the paper, some theoretical properties connecting these two types of many-valued implications are presented. Obtained results can serve as a basis for an integration of data-mining procedures discovering association rules and rule-based knowledge systems.

Bubble columns dynamics inferred from the motion of a radioactive tracer followed by axially aligned detectors

Bubble column dynamics can be properly studied with advanced velocimetry techniques, like radioactive particle tracking (RPT). However, sophisticated methods are scarcely used in the industrial environment; simpler methodologies like nuclear gauge scanning are generally preferred even if they provide more limited information. This work explores the feasibility of inferring information of the underlying dynamics in bubble columns, using a set of axially aligned detectors (AADs). The AADs scan simultaneously different column heights when a unique radioactive tracer is left free within the examined vessel, thus avoiding the cumbersome calibration required for RPT. The combined response of the AADs is analyzed by means of a data mining procedure to determine quantitative indexes related to the underlying flow regime.

The role of intelligent agents and data mining in electronic partnership management

The marketspaces of the “New Economy” and the eServices revolution have enabled the formation of new types of partnerships which are electronically mediated. Web-based electronic commerce has also brought a tremendous increase in the volume of data that can be mined for valuable managerial knowledge. The data mining procedures used in this process can be enhanced by employing intelligent agents. This paper describes emerging electronic partnerships between players in developing electronic marketspaces and identifies typical data flows between such players, with an analysis of the potential role of data mining and intelligent agent technology. By identifying the complex nature of information flows between the vast numbers of economic entities, we identify opportunities for applying data mining techniques that can lead to knowledge discovery. In particular, we show how a Generic Agent-based data Mining Architecture (GAMA) can be customized to support managerial decision-making and problem solving in a networked economy. A prototype implementation of GAMA is presented, along with a demonstration of the some of the capabilities of the system. Finally, we explore the role of agents in promoting and maintaining strong automated relationships between various strategic partners.

Mining the Shirt Sizes for Indian Men by Clustered Classification

In garment production engineering, sizing system plays an important role for manufacturing of clothing. The standards for defining the size label are a critical issue. Locating the right garment size for a customer depends on the label as an interface. In this research work intend to approach that it could be used for developing sizing systems by data mining techniques applied to Indian anthropometric dataset. We propose a new approach of two-stage data mining procedure for labelling the shirt types exclusively for Indian men. In the first stage , clustering technique applied on the original dataset, to categorise the size labels. Then these clusters are used for supervised learning in the second stage for classification. A sizing system classifies a specific population into homogeneous subgroups based on some key body dimensions. The space with these dimensions gives raise to complexity for finding uniform standards. This enables us to have an interface as a communication tool among manufacturers, retailers and consumers. This sizing system is developed for the mens age ranges between 25 and 66 years. Main attribute happens to be the chest size as clearly visible in the data set. We have obtained classifications for mens shirt attributes based on clustering techniques.

Risk prediction system for pharmacological problems

This work considers the results of laboratory investigations carried out to create a system for predicting cardiac necrosis risks that would be based on algorithms and procedures of data mining. Continuous data that indicated changes in the heartbeat and descriptive characteristics of the test animals were used. The procedures of data mining used included the selection of attributes, preprocessing, clusterization, classification, forecasting, and the data analysis. The belonging of an object to a particular group is found out during the clusterization and preprocessing of continuous data. Correlation among different descriptive characteristics of the animals is determined. The correlation between the continuous data and descriptive characteristics is found using a classification whose results are integrated in the form of conditional rules with the evaluation of the cardiac necrosis risks obtained in the laboratory. The resulted conditional rules and descriptive characteristics of the test animals provide the basis for predicting the cardiac necrosis risks.

Weed Vegetation of Sugarcane Cropping Systems of Northern Argentina: Data-Mining Methods for Assessing the Environmental and Management Effects on Species Composition

Weed composition may vary because of natural environment, management practices, and their interactions. In this study we presented a systematic approach for analyzing the relative importance of environmental and management factors on weed composition of the most conspicuous species in sugarcane. A data-mining approach represented byk-means cluster and classification and regression trees (CART) were used for analyzing the 11 most frequent weeds recorded in sugarcane cropping systems of northern Argentina. Data of weed abundance and explanatory factors contained records from 1976 sugarcane fields over 2 consecutive years. Thek-means method selected five different weed clusters. One cluster contained 44% of the data and exhibited the lowest overall weed abundance. The other four clusters were dominated by three perennial species, bermudagrass, johnsongrass, and purple nutsedge, and the annual itchgrass. The CART model was able to explain 44% of the sugarcane's weed composition variability. Four of the five clusters were represented in the terminal nodes of the final CART model. Sugarcane burning before harvesting was the first factor selected in the CART, and all nodes resulting from this split were characterized by low abundance of weeds. Regarding the predictive power of the variables, rainfall and the genotype identity were the most important predictors. These results have management implications as they indicate that the genotype identity would be a more important factor than crop age when designing sugarcane weed management. Moreover, the abiotic control of crop–weed interaction would be more related to rainfall than the environmental heterogeneity related to soil type, for example soil fertility. Although all these exploratory patterns resulting from the CART data-mining procedure should be refined, it became clear that this information may be used to develop an experimental framework to study the factors driving weed assembly.

A hybrid data mining GRASP with path-relinking

The exploration of hybrid metaheuristics—combination of metaheuristics with concepts and processes from other research areas—has been an important trend in combinatorial optimization research. An instance of this study is the hybrid version of the GRASP metaheuristic that incorporates a data mining process. Traditional GRASP is an iterative metaheuristic which returns the best solution reached over all iterations. In the hybrid GRASP proposal, after executing a significant number of iterations, the data mining process extracts patterns from an elite set of sub-optimal solutions for the optimization problem. These patterns present characteristics of near optimal solutions and can be used to guide the following GRASP iterations in the search through the combinatorial solution space. The hybrid data mining GRASP has been successfully applied for different combinatorial problems: the set packing problem, the maximum diversity problem, the server replication for reliable multicast problem and the p-median problem. In this work, we show that, not only the traditional GRASP, but also GRASP improved with the path-relinking heuristic—a memory-based intensification strategy—could benefit from exploring a data mining procedure. Computational experiments, comparing traditional GRASP with path-relinking and different path-relinking hybrid proposals, showed that employing the combination of path-relinking and data mining made the GRASP find better results in less computational time. Another contribution of this work is the application of the path-relinking hybrid proposal for the 2-path network design problem, which improved the state-of-the-art solutions for this problem.

AN INNOVATIVE WEB MINING APPLICATION ON BLOGS - A LAYOUT

Blogs and Web services agree to express user’s opinions and interests, in the form of small text messages which gives abbreviated and highly personalized remarks in real-time. Recognizing emotion is really significant for a text-based communication tool such as blogs. Nowadays, user opinions in the structure of comments, reviews in blogs have been utilized by researchers for various purposes. Among them the application of sentiment analysis techniques to these opinions is an interesting one. This paper deals with a proposal of a software structural design for constructing Web mining applications in the blog world. The design includes blog crawling and data mining algorithms, to offer a full-fledged and flexible key for constructing general-purpose Web mining applications. The structural design allocates some significant customizations, such as the construction of adapters for reading text from different blogs, and the utilization of different pre-processing methods and data mining procedures. The core of this paper is on explaining the innovative software structural design of the general framework offering thorough information about the data mining sub-framework.

Predicting gully initiation: comparing data mining techniques, analytical hierarchy processes and the topographic threshold

ABSTRACTPredicting gully initiation at catchment scale was done previously by integrating a geographical information system (GIS) with physically based models, statistical procedures or with knowledge‐based expert systems. However, the reliability and validity of applying these procedures are still questionable. In this work, a data mining (DM) procedure based on decision trees was applied to identify areas of gully initiation risk. Performance was compared with the analytic hierarchy process (AHP) expert system and with the commonly used topographic threshold (TT) technique. A spatial database was used to test the models, composed of a target variable (presence or absence of initial points) and ten independent environmental, climatic and human‐induced variables. The following findings emerged: using the same input layers, DM provided better predictive ability of gully initiation points than the application of both AHP and TT. The main difference between DM and TT was the very high overestimation inherent in TT. In addition, the minimum slope observed for soil detachment was 2°, whereas in other studies it is 3°. This could be explained by soil resistance, which is substantially lower in agricultural fields, while most studies test unploughed soil. Finally, rainfall intensity events >62.2 mm h‐1(for a period of 30 min) were found to have a significant effect on gully initiation. Copyright © 2012 John Wiley & Sons, Ltd.

Applying Randomness Effectively Based on Random Forests for Classification Task of Datasets of Insufficient Information

Random forests are known to be good for data mining of classification tasks, because random forests are robust for datasets having insufficient information possibly with some errors. But applying random forests blindly may not produce good results, and a dataset in the domain of rotogravure printing is one of such datasets. Hence, in this paper, some best classification accuracy based on clever application of random forests to predict the occurrence of cylinder bands in rotogravure printing is investigated. Since random forests could generate good results with an appropriate combination of parameters like the number of randomly selected attributes for each split and the number of trees in the forests, an effective data mining procedure considering the property of the target dataset by way of trial random forests is investigated. The effectiveness of the suggested procedure is shown by experiments with very good results.

Plantmetabolomics.org: mass spectrometry-based Arabidopsis metabolomics--database and tools update.

The PlantMetabolomics (PM) database (http://www.plantmetabolomics.org) contains comprehensive targeted and untargeted mass spectrum metabolomics data for Arabidopsis mutants across a variety of metabolomics platforms. The database allows users to generate hypotheses about the changes in metabolism for mutants with genes of unknown function. Version 2.0 of PlantMetabolomics.org currently contains data for 140 mutant lines along with the morphological data. A web-based data analysis wizard allows researchers to select preprocessing and data-mining procedures to discover differences between mutants. This community resource enables researchers to formulate models of the metabolic network of Arabidopsis and enhances the research community's ability to formulate testable hypotheses concerning gene functions. PM features new web-based tools for data-mining analysis, visualization tools and enhanced cross links to other databases. The database is publicly available. PM aims to provide a hypothesis building platform for the researchers interested in any of the mutant lines or metabolites.

Essential Attributes Identified in the Design of a Laboratory Information Management System for a High Throughput siRNA Screening Laboratory

In recent years high throughput screening operations have become a critical application in functional and translational research. Although a seemingly unmanageable amount of data is generated by these high-throughput, large-scale techniques, through careful planning, an effective Laboratory Information Management System (LIMS) can be developed and implemented in order to streamline all phases of a workflow. Just as important as data mining and analysis procedures at the end of complex processes is the tracking of individual steps of applications that generate such data. Ultimately, the use of a customized LIMS will enable users to extract meaningful results from large datasets while trusting the robustness of their assays. To illustrate the design of a custom LIMS, this practical example is provided to highlight the important aspects of the design of a LIMS to effectively modulate all aspects of an siRNA screening service. This system incorporates inventory management, control of workflow, data handling and interaction with investigators, statisticians and administrators. All these modules are regulated in a synchronous manner within the LIMS.

Gene Categories Differentially Expressed in C. elegans Age-1 Mutants of Extraordinary Longevity: New Insights From Novel Data-Mining Procedures

Two nonsense mutants of age-1, the Caenorhabditis elegans gene encoding phosphoinositide 3-kinase, live nearly 10-fold longer than wild-type controls and are exceptionally resistant to several stresses. Genome-wide expression analyses implicated downregulation of many more genes than were upregulated in second-generation age-1 homozygotes. Functional-annotation analysis, based on Gene Ontology terms, suggested that novel mechanisms may mediate the stronger phenotypes observed for these worms than with milder age-1 disruption. For the current study, the same microarray data were reanalyzed using novel meta-analytic procedures that we developed recently. First, gene p values were corrected for systematic biases based on the observed distribution for nonexpressed genes; these values were then combined to derive an aggregate p value for each functional-annotation term while adjusting for intergene covariance. This resulted in much better coverage of relevant gene categories, including many that were independently supported by other data. The number of nonredundant GO categories significantly distinguishing age-1 alleles of exceptional longevity increased from sevenfold to greater than ninefold, improving both sensitivity and specificity of selection for altered pathways and implicating previously unsuspected longevity mechanisms. Of 150 genes whose differential expression underlay significant GO terms in both comparisons, over half were up- or down-regulated in accord with longevity, whereas one third showed altered expression uniquely in the longest-lived age-1-null strains, consistent with the activation or suppression of pathways peculiar to strong age-1 mutants.

A New Approach for Handling Null Values in Web Log Using KNN and Tabu Search KNN

When the data mining procedures deals with the extraction of interesting knowledgefrom web logs is known as Web usage mining. The result of any mining issuccessful, only if the dataset under consideration is well preprocessed. One of the important preprocessing steps is handling of null/missing values. Handlings of null values have been a great bit of test for researcher. Various methods are available for estimation of null value such as k-means clustering algorithm, MARE al gorithm and fuzzy logic approach. Although all these process are notalways efficient. We propose an efficient approach for handling null value s in web log. We are using a hybrid tabu search– k nearest neighbor classifier with multiple distance function. Tabu search– KNN classifier perform feature selection of K-NN rules. We are handling null values efficiently by u sing different distance function. It is called Ensemble of function. It gives different set of feature vector. Feature selection is useful for improving the classification accuracy of NN rule. We are using different distance metric with different set of feature, so it reduces the possibility that some error will common.Therefore, proposed method is better for handling null values. The proposed method is using hybrid classifier with different distance metrics and different feature vector . It is evaluated using our MANIT database.Results have indicated that a significant increase in the performance when compared with simple K -NN classifier.

Kernel approximately harmonic projection

Dimensionality reduction is an important preprocessing procedure in computer vision, pattern recognition, information retrieval, and data mining. In this paper we present a kernel method based on approximately harmonic projection (AHP), a recently proposed linear manifold learning method that has an excellent performance in clustering. The kernel matrix implicitly maps the data into a reproducing kernel Hilbert space (RKHS) and makes the structure of data more distinct, which distributes on nonlinear manifold. It retains and extends the advantages of its linear version and keeps the sensitive to the connected components. This makes the method particularly suitable for unsupervised clustering. Besides, this method can cover various classes of nonlinearities with different kernels. We experiment the new method on several well-known data sets to demonstrate its effectiveness. The results show that the new algorithm performs a good job and outperforms other classic algorithms on those data sets.

Hybrid models in agent-based environmental decision support

Abstract Providing informational support in decision making is one of the priority directions of research in the sphere of public health management. Modern approaches suggest wide usage of intelligent data mining methods and Web-services, but just a few enable to study a complex system from an interdisciplinary point of view. In this paper an agent-based decision support system (ADSS), which embodies the principles of the interdisciplinary approach and facilitates multi-focal view and examination of the “Environment–Public health” system, is introduced. The detailed design of the system with the emphasis on the roles, scenarios and its implementation is presented. The data mining procedures used for data preparation, modeling and simulation, which include statistics, methods of artificial intelligence and hybrid models in form of cascade committee machines, are described. Then, the advantages of the proposed hybrid models against “singular” modeling methods are demonstrated. Finally, a case study for a selected region is presented and its results are discussed.

A Comparative Study of Nonlinear Machine Learning for the "In Silico" Depiction of Tyrosinase Inhibitory Activity from Molecular Structure.

In the preset report, for the first time, support vector machine (SVM), artificial neural network (ANN), Bayesian networks (BNs), k-nearest neighbor (k-NN) are applied and compared on two "in-house" datasets to describe the tyrosinase inhibitory activity from the molecular structure. The data set Data I is used for the identification of tyrosinase inhibitors (TIs) including 701 active and 728 inactive compounds. Data II consists of active chemicals for potency estimation of TIs. The 2D TOMOCOMD-CARDD atom-based quadratic indices are used as molecular descriptors. The derived models show rather encouraging results with the areas under the Receiver Operating Characteristic (AURC) curve in the test set above 0.943 and 0.846 for the Data I and Data II, respectively. Multiple comparison tests are carried out to compare the performance of the models and reveal the improvement of machine learning (ML) techniques with respect to statistical ones (see Chemometr. Intell. Lab. Syst. 2010, 104, 249). In some cases, these ameliorations are statistically significant. The tests also demostrate that k-NN, despite being a rather simple approach, presents the best behavior in both data. The obtained results suggest that the ML-based models could help to improve the virtual screening procedures and the confluence of these different techniques can increase the practicality of data mining procedures of chemical databases for the discovery of novel TIs as possible depigmenting agents.