Self-organizing Tree Algorithm Research Articles

Evaluation of potential anti-metastatic and antioxidative abilities of natural peptides derived from Tecoma stans (L.) Juss. ex Kunth in A549 cells.

BackgroundTecoma stans (L.) Juss. ex Kunth is a well-known medicinal plant found in tropical and subtropical regions. It contains a broad range of bioactive compounds that exhibit many biological effects, including antidiabetic, antibacterial, and antioxidative activities. However, the effect of natural peptides from T. stans against cancer progression and free radical production is unknown. This study aims to evaluate the cytotoxic, anti-metastatic, and antioxidative activities of natural peptides from T. stans on A549 cells.MethodsThe natural peptides were extracted from the flower of T. stans using the pressurized hot water extraction (PHWE) method, followed by size exclusion chromatography and solid-phase extraction-C18. The cytotoxic and anti-metastatic effects of natural peptides were evaluated using MTT and transwell chamber assays, respectively. The free radical scavenging activity of natural peptides was determined using ABTS, DPPH, and FRAP assays. The cells were pretreated with the IC50 dosage of natural peptides and stimulated with LPS before analyzing intracellular reactive oxygen species (ROS) and proteomics.ResultsNatural peptides induced cell toxicity at a concentration of less than 1 ng/ml and markedly reduced cell motility of A549 cells. The cells had a migration rate of less than 10% and lost their invasion ability in the treatment condition. In addition, natural peptides showed free radical scavenging activity similar to standard antioxidants and significantly decreased intracellular ROS in the LPS-induced cells. Proteomic analysis revealed 1,604 differentially expressed proteins. The self-organizing tree algorithm (SOTA) clustered the protein abundances into eleven groups. The volcano plot revealed that the cancer-promoting proteins (NCBP2, AMD, MER34, ENC1, and COA4) were down-regulated, while the secretory glycoprotein (A1BG) and ROS-reducing protein (ASB6) were up-regulated in the treatment group.ConclusionThe anti-proliferative and anti-metastatic activities of natural peptides may be attributed to the suppression of several cancer-promoting proteins. In contrast, their antioxidative activity may result from the up-regulation of ROS-reducing protein. This finding suggests that natural peptides from T. stans are viable for being the new potential anti-cancer and antioxidative agents.

Serum proteomic analysis reveals the differential dose effects of crocodile oil from Crocodylus siamensis on energy metabolism in rats.

Dietary fat composition is a potential major factor affecting energy metabolism. Crocodile oil (CO) is rich in mono- and poly-unsaturated fatty acids exhibiting anti-inflammatory and healing properties. This study investigated different levels of CO consumption on alterations and expression of proteins involved in energy metabolism in rats. Twenty-one male Sprague-Dawley rats were divided into three groups and administered sterile water (N) or different doses of CO [1% or 3% (v/w) CO] orally once daily for 8 weeks. Body weight gain, food intake, energy intake, blood lipid profiles, and serum energy-related metabolites were determined. The serum proteome was analyzed using shotgun proteomics, and the functions of several candidate proteins were classified using PANTHER software. There were no significant differences in body weight or energy intake were observed between groups. However, both CO-treated groups showed significantly decreased serum triglyceride (TG) levels (p < 0.05). Moreover, post-treatment serum TG levels in the 1%CO group were significantly lower than pre-treatment compared with other groups. The serum oxaloacetate level was also significantly higher in both CO groups than in the N group. The proteomic analysis classified 4,525 serum proteins and revealed more unique proteins involved in cellular metabolic activity in both CO-treated groups than in the N group. Self-organizing tree algorithm clustering of 295 shared differentially expressed proteins in both CO-treatment groups showed that upregulated hyper-expressed protein clusters in both CO groups were associated with catalytic activity and molecular activity on the same levels. CO simultaneously enhances energy metabolism and improves lipid profiles.

Technique of Gene Expression Profiles Extraction Based on the Complex Use of Clustering and Classification Methods.

In this paper, we present the results of the research concerning extraction of informative gene expression profiles from high-dimensional array of gene expressions considering the state of patients’ health using clustering method, ML-based binary classifiers and fuzzy inference system. Applying of the proposed stepwise procedure can allow us to extract the most informative genes taking into account both the subtypes of disease or state of the patient’s health for further reconstruction of gene regulatory networks based on the allocated genes and following simulation of the reconstructed models. We used the publicly available gene expressions data as the experimental ones which were obtained using DNA microarray experiments and contained two types of patients’ gene expression profiles—the patients with lung cancer tumor and healthy patients. The stepwise procedure of the data processing assumes the following steps—in the beginning, we reduce the number of genes by removing non-informative genes in terms of statistical criteria and Shannon entropy; then, we perform the stepwise hierarchical clustering of gene expression profiles at hierarchical levels from 1 to 10 using the SOTA (Self-Organizing Tree Algorithm) clustering algorithm with correlation distance metric. The quality of the obtained clustering was evaluated using the complex clustering quality criterion which is considered both the gene expression profiles distribution relative to center of the clusters where these gene expression profiles are allocated and the centers of the clusters distribution. The result of this stage execution was a selection of the optimal cluster at each of the hierarchical levels which corresponded to the minimum value of the quality criterion. At the next step, we have implemented a classification procedure of the examined objects using four well known binary classifiers—logistic regression, support-vector machine, decision trees and random forest classifier. The effectiveness of the appropriate technique was evaluated based on the use of ROC (Receiver Operating Characteristic) analysis using criteria, included as the components, the errors of both the first and the second kinds. The final decision concerning the extraction of the most informative subset of gene expression profiles was taken based on the use of the fuzzy inference system, the inputs of which are the results of the appropriate single classifiers operation and the output is the final solution concerning state of the patient’s health. To our mind, the implementation of the proposed stepwise procedure of the informative gene expression profiles extraction create the conditions for the increasing effectiveness of the further procedure of gene regulatory networks reconstruction and the following simulation of the reconstructed models considering the subtypes of the disease and/or state of the patient’s health.

An improved ant-based algorithm based on heaps merging and fuzzy c-means for clustering cancer gene expression data

The microarray technology enables the analysis of the gene expression data and the understanding of the important biological processes in an efficient way. We have developed an efficient clustering scheme for microarray gene expression data based on correlation-based feature selection, ant-based clustering, fuzzy c-means algorithm and a novel heaps merging heuristic. The algorithm utilizes the feature selection algorithm to overcome the high-dimensionality problem encountered in bioinformatics domain. Based on extensive empirical analysis on microarray data, clustering quality of the ant-based clustering algorithm is enhanced with the use of fuzzy c-means algorithm and heaps merging heuristic. The performance of the proposed clustering scheme is compared with k-means, PAM algorithm, CLARA, self-organizing map, hierarchical clustering, divisive analysis clustering, self-organizing tree algorithm, hybrid hierarchical clustering, consensus clustering, AntClass algorithm and fuzzy c-means clustering algorithms. The experimental results indicate that the proposed clustering scheme yields better performance in clustering cancer gene expression data.

Uterine fluid proteome changes during diapause and resumption of embryo development in roe deer (Capreolus capreolus).

The uterine microenvironment during pre-implantation presents a pro-survival milieu and is essential for embryo elongation in ruminants. The European roe deer (Careolus capreolus) pre-implantation embryo development is characterised by a 4-month period of reduced development, embryonic diapause, after which the embryo rapidly elongates and implants. We investigated the uterine fluid proteome by label-free liquid chromatography tandem mass spectrometry at four defined stages covering the phase of reduced developmental pace (early diapause, mid-diapause and late diapause) and embryo elongation. We hypothesised that embryo development during diapause is halted by the lack of signals that support progression past the blastocyst stage. Three clusters of differentially abundant proteins were identified by a self-organising tree algorithm: (1) gradual reduction over development; (2) stable abundance during diapause, followed by a sharp rise at elongation; and (3) gradual increase over development. Proteins in the different clusters were subjected to gene ontology analysis. ‘Cellular detoxification’ in cluster 1 was represented by alcohol dehydrogenase, glutathione S-transferase and peroxiredoxin-2. ATP-citrate synthase, nucleolin, lamin A/C, and purine phosphorylase as cell proliferation regulators were found in cluster 2 and ‘cortical cytoskeleton’, ‘regulation of substrate adhesion-dependent cell spreading’ and ‘melanosome’ were present in cluster 3. Cell cycle promoters were higher abundant at elongation than during diapause, and polyamines presence indicates their role in diapause regulation. This study provides a comprehensive overview of proteins in the roe deer uterine fluid during diapause and forms a basis for studies aiming at understanding the impact of the lack of cell cycle promoters during diapause.

Self-Organizing Tree Algorithm (SOTA) Clustering for Defining Level of Service (LOS) Criteria of Urban Streets

This study is intended to define the Free Flow Speed (FFS) ranges of urban street classes and speed ranges of Level of Service (LOS) categories. In order to accomplish the study FFS data and average travel speed data were collected on five urban road corridors in the city of Mumbai, India. Mid-sized vehicle (car) mounted with Global Positioning System (GPS) device was used for the collection of large number of speed data. Self-Organizing Tree Algorithm (SOTA) clustering method and five cluster validation measures were used to classify the urban streets and LOS categories. The above study divulges that the speed ranges for different LOS categories are lower than that suggested by Highway Capacity Manual (HCM) 2000. Also it has been observed that average travel speed of LOS categories expressed in percentage of free flow speeds closely resembles the percentages mentioned in HCM 2010.

The Dynamics of microRNA Transcriptome in Bovine Corpus Luteum during Its Formation, Function, and Regression.

The formation, function, and subsequent regression of the ovarian corpus luteum (CL) are dynamic processes that enable ovary cyclical activity. Studies in whole ovary tissue have found microRNAs (miRNAs) to by critical for ovary function. However, relatively little is known about the role of miRNAs in the bovine CL. Utilizing small RNA next-generation sequencing we profiled miRNA transcriptome in bovine CL during the entire physiological estrous cycle, by sampling the CL on days: d 1–2, d 3–4, and d 5–7 (early CL, eCL), d 8–12 (mid CL, mCL), d 13–16 (late CL, lCL), and d > 18 (regressed CL, rCL). We characterized patterns of miRNAs abundance and identified 42 miRNAs that were consistent significantly different expressed (DE) in the eCL relative to their expression at each of the analyzed stages (mCL, lCL, and rCL). Out of these, bta-miR-210-3p, −2898, −96, −7-5p, −183-5p, −182, and −202 showed drastic up-regulation with a fold-change of ≥2.0 and adjusted P < 0.01 in the eCL, while bta-miR-146a was downregulated at lCL and rCL vs. the eCL. Another 24, 11, and 21 miRNAs were significantly DE only between individual comparisons, eCL vs. the mCL, lCL, and rCL, respectively. Irrespective of cycle stage two miRNAs, bta-miR-21-5p and bta-miR-143 were identified as the most abundant miRNAs species and show opposing expression abundance. Whilst bta-miR-21-5p peaked in number of reads in the eCL and was significantly downregulated in the mCL and lCL, bta-miR-143 reached its peak in the rCL and is significantly downregulated in the eCL. MiRNAs with significant DE in at least one cycle stage (CL class) were further grouped into eight distinct clusters by the self-organizing tree algorithm (SOTA). Half of the clusters contain miRNAs with low-expression, whilst the other half contain miRNAs with high-expression levels during eCL. Prediction analysis for significantly DE miRNAs resulted in target genes involved with CL formation, functionalization and CL regression. This study is the most comprehensive profiling of miRNA transcriptome in bovine CL covering the entire estrous cycle and provides a compact database for further functional validation and biomarker identification relevant for CL viability and fertility.

Analysis of the Temporal Structure Evolution of Physical Systems with the Self-Organising Tree Algorithm (SOTA): Application for Validating Neural Network Systems on Adaptive Optics Data before On-Sky Implementation

Adaptive optics reconstructors are needed to remove the effects of atmospheric distortion in optical systems of large telescopes. The use of reconstructors based on neural networks has been proved successful in recent times. Some of their properties require a specific characterization. A procedure, based in time series clustering algorithms, is presented to characterize the relationship between temporal structure of inputs and outputs, through analyzing the data provided by the system. This procedure is used to compare the performance of a reconstructor based in Artificial Neural Networks, with one that shows promising results, but is still in development, in order to corroborate its suitability previously to its implementation in real applications. Also, this procedure could be applied with other physical systems that also have evolution in time.

VALIDATION OF CLUSTERING METHODS FOR MEDICAL DATA SETS

Abstract Introduction: Data mining techniques have been increasingly applied to medical data in the past decade and are divided into two categories: predictive algorithms such as classification and descriptive algorithms such as clustering and association rule mining (ARM). Clustering means partitioning a data set into a set of clusters in such a way that the samples belonging to the same clusters are similar and those belonging to different clusters are dissimilar. These algorithms are currently used as a preprocessing technique prior to medical data mining and analyzing. However, clustering algorithms have different behaviors depending on the features of the data set. Therefore, in most applications, results of clustering are evaluated in terms of some validity clustering measures. In addition, most of the medical data sets are usually complex, with nonlinear patterns, which are extremely large and difficult to cluster. The aim of this study is to performance analyze several clustering techniques based on the basis of two different classes of clustering quality measures, named internal and stability, over the medical data sets. Methods: In this study, the performance of six common clustering algorithms, including hierarchal clustering, K-means, fuzzy C-means (FCM), self-organizing tree algorithm (SOTA), Diana and partitioning around medoids (PAM), has been examined by using four different medical data sets, which are available on UCI repositories, including the Indian liver patient data set (ILPD), Pima Indians diabetes, breast cancer, and statlog heart disease. The results of clustering methods have been evaluated based on two different classes of clustering quality measures called “internal” and “stability.” Stability measures include average proportion of non-overlap (APN), average distance (AD), average distance between means (ADM), and figure of merit (FOM) and internal measures consist of connectivity, Silhouette and Dunn index. Results: Due to given data sets consisting of two classes of samples, the number of cluster is assumed to two. The evaluation results showed that hierarchical clustering is the best algorithm to cluster all of the current data sets on the basis of internal validity measures. While based on stability measures, a Diana algorithm for ILPD and breast data sets, hierarchical clustering method for heart data set and SOTA for Pima data set outperformed other clustering algorithms. Conclusion: In this study, performance of several clustering techniques on the basis of internal and stability clustering measures over the medical data sets were investigated. The hierarchical clustering and Diana methods revealed better results. The number of clusters is assumed to two, while this parameter can have an effect on performance of different clustering algorithms. Future work will mainly cover the optimal estimation number of clusters and also consider more clustering methods for evaluation.

Genome-wide identification of cassava R2R3 MYB family genes related to abscission zone separation after environmental-stress-induced abscission.

Cassava plants (Manihot esculenta Crantz) resist environmental stresses by shedding leaves in leaf pulvinus abscission zones (AZs), thus leading to adaptation to new environmental conditions. Little is known about the roles of cassava R2R3 MYB factors in regulating AZ separation. Herein, 166 cassava R2R3 MYB genes were identified. Evolutionary analysis indicated that the 166 R2R3 MYB genes could be divided into 11 subfamilies. Transcriptome analysis indicated that 26 R2R3 MYB genes were expressed in AZs across six time points during both ethylene- and water-deficit stress-induced leaf abscission. Comparative expression profile analysis of similar SOTA (Self Organizing Tree Algorithm) clusters demonstrated that 10 R2R3 MYB genes had similar expression patterns at six time points in response to both treatments. GO (Gene Ontology) annotation confirmed that all 10 R2R3 MYB genes participated in the responses to stress and ethylene and auxin stimuli. Analysis of the putative 10 R2R3 MYB promoter regions showed that those genes primarily contained ethylene- and stress-related cis-elements. The expression profiles of the genes acting downstream of the selected MYBs were confirmed to be involved in cassava abscission zone separation. All these results indicated that R2R3 MYB plays an important regulatory role in AZ separation.

Exposure to various abscission-promoting treatments suggests substantial ERF subfamily transcription factors involvement in the regulation of cassava leaf abscission.

BackgroundCassava plants (Manihot esculenta Crantz) have obvious abscission zone (AZ) structures in their leaf pulvinus-petioles. Cassava leaf abscission can be triggered by either 17 days of water-deficit stress or 4 days of ethylene treatment. To date, little is known about cassava AP2/ERF factors, and less is known regarding their roles in regulating abscission zone development.ResultsHere, the cassava and Arabidopsis AP2/ERF genes were compared, finding that the cassava genome contains approximately 1.54-fold more ERF subfamily than the Arabidopsis genome. Microarray analysis was used to identify the AP2/ERF genes that are expressed in cassava leaf pulvinus-petiole abscission zones by comparing the AP2/ERF gene expression profiles of ethylene- and water-deficit stress-induced leaf abscission. In total, 99 AP2/ERF genes were identified as expressed in AZs across six time points during both ethylene- and water-deficit stress-induced leaf abscission. Comparative expression profile analysis of similar SOTA (Self Organizing Tree Algorithm) clusters at six time points during ethylene- and water-deficit stress-induced leaf abscission demonstrated that 20 ERF subfamily genes had similar expression patterns in response to both treatments. GO (Gene Ontology) annotation confirmed that all 20 ERF subfamily genes participate in ethylene-mediated signalling. Analysis of the putative ERF promoter regions shown that the genes contained primarily ethylene- and stress-related cis-elements. Further analysis of ACC oxidase activity in AZs across six time points during abscission shown increased ethylene production in response to both ethylene and water-deficit stress; however, the difference was more dramatic for water-deficit stress. Finally, the expression ratios of 20 ERF subfamily genes were analysed in two cassava cultivars, ‘KU50’ and ‘SC5’, that exhibit different levels of leaf abscission when challenged with the same water-deficit stress. The analysis indicated that most of the ERF genes were expressed at higher levels in the precocious abscission ‘KU50’ cultivar than in the delayed abscission ‘SC5’ cultivar.ConclusionCcomparative analysis of both ethylene- and water-deficit stress-induced leaf abscission shown that the ERF subfamily functions in the regulation of cassava abscission zone development.Electronic supplementary materialThe online version of this article (doi:10.1186/s12864-016-2845-5) contains supplementary material, which is available to authorized users.

A Novel Intrusion Detection System Based on Trust Evaluation to Defend Against DDoS Attack in MANET

With the increasing demand of data communication in Internet and electronic commerce environments, security of the data is the prime concern. Large-scale collaborative wireless mobile ad hoc networks may face attacks and damages due to harsh behavior of the malicious nodes. To protect the systems from the intrusion of the attackers, security of the system has to be improvised. In researches involving the designing of the intrusion detection system (IDS), performance efficiency of the system is bound to be compromised. For an effective data communication process in the secured system, there is a need for better IDS without reducing the performance metrics. Intrusion detection is the progression of monitoring node movements and data transmission events that occur in a system for possible intrusions. Distributed denial of service (DDOS) attacks are the primary threat for security in the collaborative wireless Mobile Ad hoc networks. The attacks due to DDOS are much severe when compared to the non DDOS attacks. So proper preventive measures are necessary to detect and revoke such attacks. Our proposed approach involves trust-based evaluation wherein the intrusion detection is done using secured trust evaluation policies. In this paper, a novel IDS is designed using the trust evaluation metrics. This is used for the detection of the flooding DDOS attacks in the networked architecture. The proposed system combines the existing Firecol-based security procedures with Dynamic Growing Self-Organizing Tree Algorithm in the trust evaluation-based environment. Simulation results show that the Trust-based IDS is found to be better in terms of Security metrics viz. Detection probability and Performance metrics viz. Packet Data Ratio, Average Delay, Throughput and Energy Consumption.

Particle-Size Analysis for the Pike 1 Project, McMurray Formation

The unconsolidated sands of the Lower Cretaceous McMurray formation are the primary host of the Athabasca oil-sands deposit in Alberta, Canada. Alberta has one of the world’s largest nonconventional hydrocarbon resources with an estimated 1.8 trillion bbl of heavy oil (ERCB ST98-2013). The Pike 1 Project is a joint venture between Devon Canada Corporation (operator) and BP Canada Energy Group ULC. The Pike 1 Project is currently under a multiyear appraisal program to evaluate the McMurray bitumen resources that are amenable to steam-assisted gravity drainage(SAGD). In the Pike 1 Project area, the particle-size distribution (PSD) of the middle McMurray reservoir sands is highly variable because of the complex nature of the depositional environment. In order to understand the McMurray reservoir sands, Devon has exercised rigorous laboratory sampling and quality-control procedures to confirm the comprehensiveness of the PSD data set. By use of an unsupervised hierarchical classification technique, a dynamically growing self-organizing tree algorithm was used to cluster all of the PSD histograms from within the bitumen net-pay zone into one of four sand classes on the basis of similarity. Each sand class has a distinct PSD and permeability range. Using the sand classes, Devon extracted select intervals of core that closely matched the PSD histogram of each class to provide physical samples, termed “sandprints,” to fulfill sand-control-testing objectives. Further work included integrating the sand classes within the Pike 1 geological model, 3D permeability mapping, and upscaled sand-class volumes for SAGD well-pad optimization. This paper describes the process by which Devon has evaluated and classified the Pike 1 PSD data set into distinct sand classes within the unconsolidated middle McMurray reservoir. Devon’s methodology of acquiring these sands, necessary for sand-control testing, is also discussed in detail, emphasizing the overall effectiveness of the process. By use of this innovative PSD classification process as a supporting tool to reservoir characterization, Devon intends to realize the following benefits: • SAGD horizontal-well-pair placement optimization • A more methodical approach to laboratory testing of horizontal-liner technology for SAGD producers and injectors • Improved reservoir management

Genomic and proteomic analysis with dynamically growing self organising tree (DGSOT) for measuring clinical outcomes of cancer

Genomics and proteomics microarray technologies are used for analysing molecular and cellular expressions of cancer. This creates a challenge for analysis and interpretation of the data generated as it is produced in large volumes. The current review describes a combined system for genetic, molecular interpretation and analysis of genomics and proteomics technologies that offers a wide range of interpreted results. Artificial neural network systems technology has the type of programmes to best deal with these large volumes of analytical data. The artificial system to be recommended here is to be determined from the analysis and selection of the best of different available technologies currently being used or reviewed for microarray data analysis. The system proposed here is a tree structure, a new hierarchical clustering algorithm called a dynamically growing self-organizing tree (DGSOT) algorithm, which overcomes drawbacks of traditional hierarchical clustering algorithms. The DGSOT algorithm combines horizontal and vertical growth to construct a mutlifurcating hierarchical tree from top to bottom to cluster the data. They are designed to combine the strengths of Neural Networks (NN), which have speed and robustness to noise, and hierarchical clustering tree structure which are minimum prior requirement for number of clusters specification and training in order to output results of interpretable biological context. The combined system will generate an output of biological interpretation of expression profiles associated with diagnosis of disease (including early detection, molecular classification and staging), metastasis (spread of the disease to non-adjacent organs and/or tissues), prognosis (predicting clinical outcome) and response to treatment; it also gives possible therapeutic options ranking them according to their benefits for the patient. Key words : Genomics, proteomics, microarray, dynamically growing self-organizing tree (DGSOT).

Scene-Oriented Hierarchical Classification of Blurry and Noisy Images

A system for scene-oriented hierarchical classification of blurry and noisy images is proposed. It attempts to simulate important features of the human visual perception. The underlying approach is based on three strategies: extraction of essential signatures captured from a global context, simulating the global pathway; highlight detection based on local conspicuous features of the reconstructed image, simulating the local pathway; and hierarchical classification of extracted features using probabilistic techniques. The techniques involved in hierarchical classification use input from both the local and global pathways. Visual context is exploited by a combination of Gabor filtering with the principal component analysis. In parallel, a pseudo-restoration process is applied together with an affine invariant approach to improve the accuracy in the detection of local conspicuous features. Subsequently, the local conspicuous features and the global essential signature are combined and clustered by a Monte Carlo approach. Finally, clustered features are fed to a self-organizing tree algorithm to generate the final hierarchical classification results. Selected representative results of a comprehensive experimental evaluation validate the proposed system.

Network Based Anomaly Intrusion Detection System Using SVM

The security and integrity of a computer system is compromised when an intrusion occurs. It becomes impossible for legitimate users to access different network services when network-based attacks purposely occupy or sabotage network resources and services. Our proposed method is a scalable detection method for network based anomalies. We use Support Vector Machines (SVM) for classification. This paper presents a method for enhancing the training time of SVM, particularly when dealing with large data sets, using hierarchical clustering technique. We use the Dynamically Growing Self-Organizing Tree (DGSOT) algorithm for clustering because it has proved to overcome the problems of traditional hierarchical clustering algorithms (e.g., hierarchical agglomerative clustering). Clustering analysis helps to find the boundary points, which are the most qualified data points to train SVM, between any two classes. We present a new approach of combination of SVM and DGSOT, which begins with an initial training set and expands it gradually using the clustering structure produced by the DGSOT algorithm. We show that our proposed variations contribute significantly in improving the training process of SVM with high percentage of detection accuracy.

Network based anomaly intrusion detection system using SVM

The security and integrity of a computer system is compromised when an intrusion occurs. It becomes impossible for legitimate users to access different network services when network-based attacks purposely occupy or sabotage network resources and services. Our proposed method is a scalable detection method for network based anomalies. We use Support Vector Machines (SVM) for classification. This paper presents a method for enhancing the training time of SVM, particularly when dealing with large data sets, using hierarchical clustering technique. We use the Dynamically Growing Self-Organizing Tree (DGSOT) algorithm for clustering because it has proved to overcome the problems of traditional hierarchical clustering algorithms (e.g., hierarchical agglomerative clustering). Clustering analysis helps to find the boundary points, which are the most qualified data points to train SVM, between any two classes. We present a new approach of combination of SVM and DGSOT, which begins with an initial training set and expands it gradually using the clustering structure produced by the DGSOT algorithm. We show that our proposed variations contribute significantly in improving the training process of SVM with high percentage of detection accuracy. Keywords: SVM, classification, Intrusion detection, Intrusion detection System, Network Security

Transcriptome analysis of a spontaneous mutant in sweet orange [Citrus sinensis (L.) Osbeck] during fruit development

Bud mutations often arise in citrus. The selection of mutants is one of the most important breeding channels in citrus. However, the molecular basis of bud mutation has rarely been studied. To identify differentially expressed genes in a spontaneous sweet orange [C. sinensis (L.) Osbeck] bud mutation which causes lycopene accumulation, low citric acid, and high sucrose in fruit, suppression subtractive hybridization and microarray analysis were performed to decipher this bud mutation during fruit development. After sequencing of the differentially expressed clones, a total of 267 non-redundant transcripts were obtained and 182 (68.2%) of them shared homology (E-value ≤1×10−10) with known gene products. Few genes were constitutively up- or down-regulated (fold change ≥2) in the bud mutation during fruit development. Self-organizing tree algorithm analysis results showed that 95.1% of the differentially expressed genes were extensively coordinated with the initiation of lycopene accumulation. Metabolic process, cellular process, establishment of localization, response to stimulus, and biological regulation-related transcripts were among the most regulated genes. These genes were involved in many biological processes such as organic acid metabolism, lipid metabolism, transport, and pyruvate metabolism, etc. Moreover, 13 genes which were differentially regulated at 170 d after flowering shared homology with previously described signal transduction or transcription factors. The information generated in this study provides new clues to aid in the understanding of bud mutation in citrus.

Clustering of gene expression data: performance and similarity analysis

BackgroundDNA Microarray technology is an innovative methodology in experimental molecular biology, which has produced huge amounts of valuable data in the profile of gene expression. Many clustering algorithms have been proposed to analyze gene expression data, but little guidance is available to help choose among them. The evaluation of feasible and applicable clustering algorithms is becoming an important issue in today's bioinformatics research.ResultsIn this paper we first experimentally study three major clustering algorithms: Hierarchical Clustering (HC), Self-Organizing Map (SOM), and Self Organizing Tree Algorithm (SOTA) using Yeast Saccharomyces cerevisiae gene expression data, and compare their performance. We then introduce Cluster Diff, a new data mining tool, to conduct the similarity analysis of clusters generated by different algorithms. The performance study shows that SOTA is more efficient than SOM while HC is the least efficient. The results of similarity analysis show that when given a target cluster, the Cluster Diff can efficiently determine the closest match from a set of clusters. Therefore, it is an effective approach for evaluating different clustering algorithms.ConclusionHC methods allow a visual, convenient representation of genes. However, they are neither robust nor efficient. The SOM is more robust against noise. A disadvantage of SOM is that the number of clusters has to be fixed beforehand. The SOTA combines the advantages of both hierarchical and SOM clustering. It allows a visual representation of the clusters and their structure and is not sensitive to noises. The SOTA is also more flexible than the other two clustering methods. By using our data mining tool, Cluster Diff, it is possible to analyze the similarity of clusters generated by different algorithms and thereby enable comparisons of different clustering methods.

A new intrusion detection system using support vector machines and hierarchical clustering

Whenever an intrusion occurs, the security and value of a computer system is compromised. Network-based attacks make it difficult for legitimate users to access various network services by purposely occupying or sabotaging network resources and services. This can be done by sending large amounts of network traffic, exploiting well-known faults in networking services, and by overloading network hosts. Intrusion Detection attempts to detect computer attacks by examining various data records observed in processes on the network and it is split into two groups, anomaly detection systems and misuse detection systems. Anomaly detection is an attempt to search for malicious behavior that deviates from established normal patterns. Misuse detection is used to identify intrusions that match known attack scenarios. Our interest here is in anomaly detection and our proposed method is a scalable solution for detecting network-based anomalies. We use Support Vector Machines (SVM) for classification. The SVM is one of the most successful classification algorithms in the data mining area, but its long training time limits its use. This paper presents a study for enhancing the training time of SVM, specifically when dealing with large data sets, using hierarchical clustering analysis. We use the Dynamically Growing Self-Organizing Tree (DGSOT) algorithm for clustering because it has proved to overcome the drawbacks of traditional hierarchical clustering algorithms (e.g., hierarchical agglomerative clustering). Clustering analysis helps find the boundary points, which are the most qualified data points to train SVM, between two classes. We present a new approach of combination of SVM and DGSOT, which starts with an initial training set and expands it gradually using the clustering structure produced by the DGSOT algorithm. We compare our approach with the Rocchio Bundling technique and random selection in terms of accuracy loss and training time gain using a single benchmark real data set. We show that our proposed variations contribute significantly in improving the training process of SVM with high generalization accuracy and outperform the Rocchio Bundling technique.