Normal Dataset Research Articles

Feature Analysis, Evaluation and Comparisons of Classification Algorithms Based on Noisy Intrusion Dataset

Various studies have been carried on an Intrusion Detection System (IDS) environment bycomparingthe performance of various Machine Learning (ML)based on a refined intrusion dataset with an error-free environment. However, the real-world network data deals with a large amount of noisy information on transmission, and the IDS have to work in such an environment frequently. Dealing with such noisy data is, therefore, a challenging issue in an IDS environment for detecting threads from network activities. In this paper, various Data Mining (DM) and ML algorithms are evaluated and compared by normal and noisy dataset prepared from KDD’99 and NSL-KDD dataset (10%-20% Noise). The empirical results demonstrate that NN (SOM) is far better compared to other tested algorithms regarding robustness tonoisy environment; however,JRip and J48 from the tree family outperform others regarding overall performance matrices. Feature dependency on datasets for a specific classifier is analyzedby Performance-based Method of Ranking (PMR). The evaluation results statistically proved that each classifier has a unique combination of a feature subset to results optimal performance. Empirical results demonstrate that evaluations of IDS based on NSL-KDD give more realistic results compared to theKDD’99 original dataset.

An Efficient and Scalable Density-based Clustering Algorithm for Normalize Data

Data clustering is a method of putting same data object into group. A clustering rule does partitions of a data set into many groups supported the principle of maximizing the intra-class similarity and minimizing the inter-class similarity. Finding clusters in object, particularly high dimensional object, is difficult when the clusters are different shapes, sizes, and densities, and when data contains noise and outliers. This paper provides a new clustering algorithm for normalized data set and proven that our new planned clustering approach work efficiently when dataset are normalized.

Skeleton clustering by multi-robot monitoring for fall risk discovery

This paper tackles the problem of discovering subtle fall risks using skeleton clustering by multi-robot monitoring. We aim to identify whether a gait has fall risks and obtain useful information in inspecting fall risks. We employ clustering of walking postures and propose a similarity of two datasets with respect to the clusters. When a gait has fall risks, the similarity between the gait which is being observed and a normal gait which was monitored in advance exhibits a low value. In subtle fall risk discovery, unsafe skeletons, postures in which fall risks appear slightly as instabilities, are similar to safe skeletons and this fact causes the difficulty in clustering. To circumvent this difficulty, we propose two instability features, the horizontal deviation of the upper and lower bodies and the curvature of the back, which are sensitive to instabilities and a data preprocessing method which increases the ability to discriminate safe and unsafe skeletons. To evaluate our method, we prepare seven kinds of gait datasets of four persons. To identify whether a gait has fall risks, the first and second experiments use normal gait datasets of the same person and another person, respectively. The third experiments consider that how many skeletons are necessary to identify whether a gait has fall risks and then we inspect the obtained clusters. In clustering more than 500 skeletons, the combination of the proposed features and our preprocessing method discriminates gaits with fall risks and without fall risks and gathers unsafe skeletons into a few clusters.

Identification of candidate genes for prostate cancer-risk SNPs utilizing a normal prostate tissue eQTL data set.

Multiple studies have identified loci associated with the risk of developing prostate cancer but the associated genes are not well studied. Here we create a normal prostate tissue-specific eQTL data set and apply this data set to previously identified prostate cancer (PrCa)-risk SNPs in an effort to identify candidate target genes. The eQTL data set is constructed by the genotyping and RNA sequencing of 471 samples. We focus on 146 PrCa-risk SNPs, including all SNPs in linkage disequilibrium with each risk SNP, resulting in 100 unique risk intervals. We analyse cis-acting associations where the transcript is located within 2 Mb (±1 Mb) of the risk SNP interval. Of all SNP–gene combinations tested, 41.7% of SNPs demonstrate a significant eQTL signal after adjustment for sample histology and 14 expression principal component covariates. Of the 100 PrCa-risk intervals, 51 have a significant eQTL signal and these are associated with 88 genes. This study provides a rich resource to study biological mechanisms underlying genetic risk to PrCa.

A Novel Graph Decomposition Approach to the Automatic Processing of Poorly Formalized Data: Innovative Ideas – A Management Case Study

Introduction In most cases, automatic processing can only be conducted on a structured and normalized data set. If the data set is unstructured, then most of data processing algorithms will fail to provide adequate results. The selection of a formalized model for the data is a process that is conducted manually. The main reason behind using manual approach is that even the state-of-the-art text mining algorithms are incapable of processing complex semantics of non-formalized and unstructured textual data. On the other hand, important patterns can be extracted only from the structured data. There are varieties of methods that provide basic data cleansing and structuring services: clusterization algorithms, normalization techniques, data cleansing, etc. A general automatic formalized model construction can be realized for more complex cases. This becomes possible due to high abstraction power of graph theory. However, the graph theory is to be used carefully--trying to apply this theory without considering the research field features can lead to models that lack needed level of details for further processing. Common approaches to the processing of connected and poorly structured data are based on graphs theory. Poorly formalized and unstructured data is one of the largest segments in the data processing. Experts agree that almost 80 to 85 percent of business-relevant information originates in unstructured form. The manual processing of unstructured data is costly and time-consuming. To interpret unstructured information such techniques as natural language processing (NLP), data mining, text analytics are used. Further, the patterns found can be organized and structured using mathematical graphs. Different kinds of graphs can be used to represent different features of data. The field of innovation management was selected to apply the results of the research, because the development of this novel approach to data structuration is a part of work on universities' innovation life cycle model (ILCM) and innovation management system. The basic components of innovation management are ideas and innovations. For example, the idea can be represented by unstructured textual descriptions and a group of illustrations. It is important to classify the idea based on its description. The manual classification becomes difficult in case of large number of ideas. The idea can be provided in the form of terms-rich description. Such description contains specific key words. There are many techniques to identify such words and phrases, however these approaches are leaving aside the problems of textual structure and semantics. There are other attributes that must be used in order to structure ideas. Considering all these attributes, the process of ideas management becomes fairly complicated, thus automatic tools become highly important for efficient innovations management. Back in 1986, Andrew H. Van de Ven has outlined four central problems in innovation management (Van de Ven, 1986): managing attention; managing ideas into good currency; managing part-whole relationships; institutional leadership. According to Van de Ven's 2004 paper, these cornerstone problems remain intact and continue to heavily influence the performance of innovation management process (Van de Ven and Engleman, 2004). In our work we will concentrate on ideas and innovations automatic structuration tasks which are in the scope of outlined problem of managing ideas into good currency. One of the major parts of this problem is to correctly define the most promising and optimal (in respect to given resources) idea. One of the best ways to estimate an idea would be to compare it with implemented analogues. The estimation can become rather complicated when considering ideas interconnections and dependencies on each other or on specific innovation. In the second section of the paper we will provide a literature review of automatic and semi-automatic graph-based decision-support approaches dealing with the processing of poorly formalized data as well as a basic review of innovation management literature. …

Beyond the one-locus-one-miRNA paradigm: microRNA isoforms enable deeper insights into breast cancer heterogeneity.

Here we describe our study of miRNA isoforms (isomiRs) in breast cancer (BRCA) and normal breast data sets from the Cancer Genome Atlas (TCGA) repository. We report that the full isomiR profiles, from both known and novel human-specific miRNA loci, are particularly rich in information and can distinguish tumor from normal tissue much better than the archetype miRNAs. IsomiR expression is also dependent on the patient's race, exemplified by miR-183-5p, several isomiRs of which are upregulated in triple negative BRCA in white but not black women. Additionally, we find that an isomiR's 5′ endpoint and length, but not the genomic origin, are key determinants of the regulation of its expression. Overexpression of distinct miR-183-5p isomiRs in MDA-MB-231 cells followed by microarray analysis revealed that each isomiR has a distinct impact on the cellular transcriptome. Parallel integrative analysis of mRNA expression from BRCA data sets of the TCGA repository demonstrated that isomiRs can distinguish between the luminal A and luminal B subtypes and explain in more depth the molecular differences between them than the archetype molecules. In conclusion, our findings provide evidence that post-transcriptional studies of BRCA will benefit from transcending the one-locus-one-miRNA paradigm and taking into account all isoforms from each miRNA locus as well as the patient's race.

An efficient intelligent analysis system for confocal corneal endothelium images

A confocal microscope provides a sequence of images of the corneal layers and structures at different depths from which medical clinicians can extract clinical information on the state of health of the patient's cornea. A hybrid model based on snake and particle swarm optimisation (S-PSO) is proposed in this paper to analyse the confocal endothelium images. The proposed system is able to pre-process images (including quality enhancement and noise reduction), detect cells, measure cell densities and identify abnormalities in the analysed data sets. Three normal corneal data sets acquired using a confocal microscope, and three abnormal confocal endothelium images associated with diseases have been investigated in the proposed system. Promising results are presented and the performance of this system is compared with manual and two morphological based approaches. The average differences between the manual and the automatic cell densities calculated using S-PSO and two other morphological based approaches is 5%, 7% and 13% respectively. The developed system will be deployable as a clinical tool to underpin the expertise of ophthalmologists in analysing confocal corneal images.

Patchy-distributed ciliate (Protozoa) diversity of eight polar communities as determined by 454 amplicon pyrosequencing

To determine ciliate diversity and community structure in the polar ecosystem efficiently, we applied the pyrosequencing technique to the polar samples. To select the appropriate sequencing depth using a ciliate-specific primer, we evaluated different pyrosequencing depths, ranging 4149–112,306 reads. At a 3% distance cutoff for clustering, 750 operational taxonomic units (OTUs) were identified, and 332 were composed of a single read (singletons). The singletons showed a 1.8-fold increase in OTU richness, although their beta diversity showed no significant changes. The ratio of singletons in each sequencing depth was sharply decreased after reaching the sequencing depth of approximately 10,000 reads, and the singletons did not completely disappear even at 73,435 qualified reads. The data set without singletons showed saturated trends in rarefaction curves. In addition, we built a normalized data set without the singletons (1227 reads × eight samples). Among the samples, one brackish water having a broad range of salinity (3–23 PSU) at the Arctic coast presented the highest OTU richness (103), while a temporal pool at the Antarctic coast with a high salinity of 53.4 PSU, showed a relatively lower OTUs (8). Each normalized sample showed a distinct community structure. Interestingly, a freshwater lake on King George Island shared relatively higher OTUs with salt-water samples (72.1%), suggesting a higher inter-relationship with closely located coastal water environment.

Abstract 1077: Use of the SVClassify algorithm to classify pediatric solid tumor translocation variant calls as likely true or false positives

Abstract With the advent of next-generation sequencing technologies, a variety of structural variant (SV) calling algorithms have been developed. However, these algorithms are generally not sensitive and specific, in part because they are biased toward identification of specific types or lengths of SVs. Thus, concordance among algorithms is often very low. At the National Institute of Standards and Technology (NIST), we have previously rigorously characterized sequencing biases across multiple sequencing platforms for the candidate NIST reference material, RM 8398 (the genome of individual NA12878). This yielded a high-confidence set of SNP and small indel (<40 bp) variant calls. To extend our methods to SVs, we have developed SVClassify, which classifies SVs as likely true or false positive variants by combining evidence from one or more sequencing datasets. For NA12878, we were able to separate a set of validated deletions from random genomic regions with false positive and false negative rates less than 5%. We also found that the set of validated deletions clustered into different categories, including heterozygous Alu deletions, homozygous Alu deletions, and other heterozygous deletions. Until now, SVClassify has only been used for classification of large deletions on the non-cancerous genome, NA12878. In order to assess the sensitivity and specificity of SVClassify for correctly classifying translocations, we are using the RSVSim R package to simulate different numbers of translocations within repeat regions of the human genome. Furthermore, we will investigate the performance of SVClassify on cancer genomes, particularly pediatric solid tumors, which exhibit extensively rearranged genomes compared to their normal counterparts. To do so, pediatric tumor and normal datasets from multiple sequencing technologies will be integrated. To represent a spectrum of translocations, we will assess variant calls from cancer genomes with varying degrees of rearrangements: neuroblastoma, Ewing sarcoma, and osteosarcoma. Finally, we will use SVclassify to classify candidate SV calls made using the Complete Genomics pipeline, as well as those made on Illumina datasets using the recently developed SMUFIN algorithm. Citation Format: Jo Lynne Harenza, Hemang M. Parikh, Jun S. Wei, Xinyu Wen, Sivasish Sindiri, Rajesh Patidar, Marc Salit, Paul S. Meltzer, Javed Khan, Justin Zook. Use of the SVClassify algorithm to classify pediatric solid tumor translocation variant calls as likely true or false positives. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 1077. doi:10.1158/1538-7445.AM2015-1077

Abstract 1321: The use of proteomics to analyze whole tumors and identify unique immuno-oncology targets for antibody-based therapeutics

Abstract The recent clinical success of the mAb therapeutics targeting immune checkpoint inhibitor proteins (PD-1/PD-L1, CTLA-4) has led to an increased appreciation of the potential of utilizing the immune system in oncology. There are two major strategies to elicit either a novel immune anti-tumor response or to reactivate a pre-existing anti-tumor response: by releasing a checkpoint inhibitory pathway via cell surface receptors (such as PD-1/PD-L1, CTLA-4) or by activation of co-stimulatory receptors (such as CD40, OX40, or GITR). Both of these strategies of immune modulation utilize cell surface receptors, and the targeting of antibody therapeutics with the appropriate functional activity to those receptors, to modify immune cell responses and allow for anti-tumor activity. The identification of novel immune-modulatory receptors with the potential to be immune-oncology therapeutic targets could be of high value to this anti-tumor approach. OGAP is a unique proteomic database that integrates information at the tissue, disease and protein isoform level across diseases, indications, and normal tissues to clarify membrane protein expression levels and profiles. Specifically, it currently holds information on ∼16,000 human proteins sequenced, ∼7,000 membrane proteins, ∼35 tissues/organs, and ∼17 cancers. OGAP is fed by a proprietary sample preparation and processing workflow that relies on state-of-the-art high-throughput mass spectrometry and data processing to provide quantitative information on over 4,000 membrane-enriched proteins. OGAP has been used to identify novel oncology therapeutic targets for both ADC and BiTE-like approaches. Utilizing OGAP, membrane proteins present in tumors from five cancer indications (Pancreatic, Lung, Breast, Colorectal and Esophageal cancer) and multiple normal tissues or cells were analyzed. Validated immune cell markers (such as CD8, OX40, CD79B, TLR1, TLR2, TLR4, TLR7, CD56, CD204 and CD207) were profiled across different normal and tumor proteomic data sets. This analysis demonstrated we detect key immune cell markers in tumors and that different immune cell populations are found in tumors from the same or different cancer indications. The proteomic data sets were next analyzed for the presence of validated immuno-oncology targets (TIM3, PDL-1 and B7-H3). The expression patterns of these immune-oncology targets were analyzed to try and identify a unique protein signature. Using this protein expression profiling approach we identified most known immune-oncology targets, validating this as an approach to identify a pool of candidate novel immune-oncology targets. To further develop our approach we also used sequence based homology searching of uncharacterized membrane proteins to improve the quality of the pool of candidate novel immune-oncology targets. Several potential novel immune-oncology targets will be presented. Citation Format: Jim Ackroyd, Arnima Bisht, Jason Allen, Lindsey Hudson, Martin Barnes, Christian Rohlff, Keith Wilson, Robert Boyd, Dee Aud. The use of proteomics to analyze whole tumors and identify unique immuno-oncology targets for antibody-based therapeutics. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 1321. doi:10.1158/1538-7445.AM2015-1321

Medical image classification based on artificial intelligence approaches: A practical study on normal and abnormal confocal corneal images

Corneal images can be acquired using confocal microscopes which provide detailed views of the different layers inside a human cornea. Some corneal problems and diseases can occur in one or more of the main corneal layers: the epithelium, stroma and endothelium. Consequently, for automatically extracting clinical information associated with corneal diseases, identifying abnormality or evaluating the normal cornea, it is important to be able to automatically recognise these layers reliably. Artificial intelligence (AI) approaches can provide improved accuracy over the conventional processing techniques and save a useful amount of time over the manual analysis time required by clinical experts. Artificial neural networks (ANNs), adaptive neuro fuzzy inference systems (ANFIS) and a committee machine (CM) have been investigated and tested to improve the recognition accuracy of the main corneal layers and identify abnormality in these layers. The performance of the CM, formed from ANN and ANFIS, achieves an accuracy of 100% for some classes in the processed data sets. Three normal corneal data sets and seven abnormal corneal images associated with diseases in the main corneal layers have been investigated with the proposed system. Statistical analysis for these data sets is performed to track any change in the processed images. This system is able to pre-process (quality enhancement, noise removal), classify corneal images, identify abnormalities in the analysed data sets and visualise corneal stroma images as well as each individual keratocyte cell in a 3D volume for further clinical analysis.

Human Cumulative Irritation Tests of Common Preservatives Used in Personal Care Products: A Retrospective Analysis of Over 45 000 Subjects.

The cumulative irritation test (CIT) is an accepted method used to evaluate the skin irritation potential and safety of individual ingredients and formulas of leave-on skin care and cosmetic compounds. Here, we report the results of CITs collected by JOHNSON & JOHNSON Consumer Companies, Inc. (Skillman, NJ), part of an extensive tiered program to evaluate product safety. In the CIT, test formulations were applied to the skin of adults (18-70 years) with no known skin disease or allergies, 3 times per week for 2 weeks using semi-occlusive clinical patches. Preservatives were 1 of up to 16 components of test formulas, and included ethylenediaminetetraacetic acid, diazolidinyl urea, 1,3-Bis(hydroxymethyl)-5,5-dimethylimidazolidine-2,4-dione, parabens, isothiazolinone, phenoxyethanol, sorbates, or benzoates. Skin sites were scored after each patch removal using a 5-point scale, with 0 = no visible reaction and 4 = erythema, marked edema, or substantial vesiculation. Scores were reported as percentage of maximal irritation score. Data were analyzed from 1363 CIT studies (over 45 000 subjects). There were no significant differences in percentage of maximal scores between formulas grouped by preservative types (p > .1). Median score across the entire dataset was 0.44, with most formulas showing none or mild irritation. Although seasonal variations were observed, no correlation was noted between score and preservative concentration. In conclusion, in a large, normal subject dataset, preservatives at typical in-use concentrations did not appear to contribute to skin irritation.

Detection and Implementation of Web-based Attacks using Attribute Length Method

With the increasing demand of web-based applications, they have become more prone to be exploited by the attackers. The purpose of this paper is to study the effects of web-based attacks and analyze the log files generated during the attacks. We have implemented Attribute Length Method proposed by Krugel for the detection of web-based attacks. In the implementation of the Attribute Length method, two different phases are used in our system i.e., learning and detection phase. In the learning phase, our implementation in Java trains the normal dataset and calculates the threshold probability value which is used in the Detection phase for the estimation of web-based attacks. In order to estimate the performance of attribute length method, we have used a log file having three different web attacks, i.e. Cross Site Scripting attack, Path Traversal attack, and Buffer Overflow attack. This method is more effective as we have considered the parameters as fixedsize tokens.

Systematic transcriptome analysis reveals tumor-specific isoforms for ovarian cancer diagnosis and therapy

Tumor-specific molecules are needed across diverse areas of oncology for use in early detection, diagnosis, prognosis and therapy. Large and growing public databases of transcriptome sequencing data (RNA-seq) derived from tumors and normal tissues hold the potential of yielding tumor-specific molecules, but because the data are new they have not been fully explored for this purpose. We have developed custom bioinformatic algorithms and used them with 296 high-grade serous ovarian (HGS-OvCa) tumor and 1,839 normal RNA-seq datasets to identify mRNA isoforms with tumor-specific expression. We rank prioritized isoforms by likelihood of being expressed in HGS-OvCa tumors and not in normal tissues and analyzed 671 top-ranked isoforms by high-throughput RT-qPCR. Six of these isoforms were expressed in a majority of the 12 tumors examined but not in 18 normal tissues. An additional 11 were expressed in most tumors and only one normal tissue, which in most cases was fallopian or colon. Of the 671 isoforms, the topmost 5% (n = 33) ranked based on having tumor-specific or highly restricted normal tissue expression by RT-qPCR analysis are enriched for oncogenic, stem cell/cancer stem cell, and early development loci--including ETV4, FOXM1, LSR, CD9, RAB11FIP4, and FGFRL1. Many of the 33 isoforms are predicted to encode proteins with unique amino acid sequences, which would allow them to be specifically targeted for one or more therapeutic strategies--including monoclonal antibodies and T-cell-based vaccines. The systematic process described herein is readily and rapidly applicable to the more than 30 additional tumor types for which sufficient amounts of RNA-seq already exist.

Database Transformation to Build Dataset for Generation of Decision Tree and Extended ER Model

In Data mining project most of the time consuming task is to prepare a required data set for data mining analysis because in general the relational database has collection of tables and views that must be joined, aggregated and transformed in order to build the required data set. As result, most of the complex SQL queries are written multiple times independently from each other and in a disorganized manner. Therefore, the database grows with many tables and views that are not present as entities in the ER model. Similarly existing SQL aggregations having some limitations to prepare normalized data sets because they return only one column per aggregated group. To address this issue, we propose simple methods to generate SQL code to return aggregated columns in a horizontal tabular layout, where every row corresponds to an observation and every column is associated to a one variable. This new class of functions is called horizontal aggregations. Horizontal aggregations is extension of standard SQL aggregation for building data sets with a horizontal denormalized layout, which is input for most of the data mining algorithms. By providing these standard normalized data-set as an input to the Decision tree generation algorithm for generating Decision tree, similarly we can generate extended ER model.

Ensemble Supervised Classification Method Using the Regions of Interest and Grey Level Co-Occurrence Matrices Features for Mammograms Data.

Background:Breast cancer is one of the most encountered cancers in women. Detection and classification of the cancer into malignant or benign is one of the challenging fields of the pathology.Objectives:Our aim was to classify the mammogram data into normal and abnormal by ensemble classification method.Patients and Methods:In this method, we first extract texture features from cancerous and normal breasts, using the Gray-Level Co-occurrence Matrices (GLCM) method. To obtain better results, we select a region of breast with high probability of cancer occurrence before feature extraction. After features extraction, we use the maximum difference method to select the features that have predominant difference between normal and abnormal data sets. Six selected features served as the classifying tool for classification purpose by the proposed ensemble supervised algorithm. For classification, the data were first classified by three supervised classifiers, and then by simple voting policy, we finalized the classification process.Results:After classification with the ensemble supervised algorithm, the performance of the proposed method was evaluated by perfect test method, which gave the sensitivity and specificity of 96.66% and 97.50%, respectively.Conclusions:In this study, we proposed a new computer aided diagnostic tool for the detection and classification of breast cancer. The obtained results showed that the proposed method is more reliable in diagnostic to assist the radiologists in the detection of abnormal data and to improve the diagnostic accuracy.

Variable moving windows based non‐Gaussian dissimilarity analysis technique for batch processes fault detection and diagnosis

In this paper, a novel variable moving windows based non‐Gaussian dissimilarity analysis technique is developed to handle the challenges related to the batch processes such as non‐linearity, non‐Gaussianity and time‐varying dynamics. First, the independent component analysis (ICA) models are developed on the normal reference data sets and the monitored data sets to extract the dominant independent component subspaces through a variable moving windows strategy. Then, non‐Gaussian dissimilarity indices are computed to evaluate the statistical independency of the extracted IC subspaces at each specific time interval. Thus, the process non‐Gaussian features are fully captured and the process trajectories distribution information of batch‐to‐batch is quantitatively estimated for online process monitoring. Further, the non‐Gaussian contribution index based on the mutual information is introduced to identify the variables that may be responsible for the process abnormality. The reliability and validity of the proposed method are verified on the fed‐batch Penicillin Fermentation process. The application results present superior fault detection and diagnosis performance compared with the PCA dissimilarity, MPCA and MICA approaches.

Fault Detection Based on Diffusion Maps and <i>k</i> Nearest Neighbor Diffusion Distance of Feature Space

Dimensionality reduction is a fundamental task of high-dimensional data analysis in order to reduce redundant information of the collected data. In this paper, we apply diffusion maps framework to address this problem, and then propose a novel fault detection technique based on the k nearest neighbor diffusion distance method of feature space. First, normal high-dimensional data sets are mapped into a low-dimensional feature space by analyzing the insightful relationship between data points, and feature space can represent major information of raw data. Subsequently, like the traditional kNN method, the sum of k nearest neighbor diffusion distance is computed and the kernel density estimation method is used to set a threshold of a normal process. Comparing these two methods, modeling using k nearest neighbor diffusion distance method of feature space can economize storage space and increase the speed of fault detection. In addition, this method can solve nonlinear equation of industrial process data, and the non-Gaussian characteristics of modeling data can be solved by using kernel density estimation method. Finally, the effectiveness of diffusion maps algorithm in the aspect of data classification is verified by the numerical examples and the superiority of the proposed method is illustrated by the monitoring of TE process.

ICA-Multiclass SVM as a Monitoring System of Complex Processes

Nowadays due to advancement of technology, the processes have become increasingly complex. So perform eectively monitor and correct diagnosis of these, has become a challenging task. this because the current processes have in common presence of combinations and correlation between analog and digital variables, presence of noise, etc. This issued is overcome by multivariate statistical methods which can eciently represent the dierent correlations between variables. One of the most popular methods for process monitoring is the Hotelling's T 2 distance but a large amount of data makes the process monitoring a dicult task. Principal Component Analysis (PCA) is used for the data reduction which means the extraction of few numbers representing the most variance of the analyzed data. However the use of these two methods is limited to normal multivariate data set and in industry rarely normal data set is presented, hence dierent methods are used for this purpose. Independent component analysis (ICA) has been used in no-normal mul- tivariate processes. In this paper a monitoring and fault diagnosis system for complex multivariate processes is presented. The new proposal is based on Multiclass Support Vector Machines (MCSVM) and the case study was addressed to the automotive industry. A comparison against a similar technique is carried out. Simulation of the new proposal shows promisory results.

The Adhesion Molecule “G-Protein Coupled Receptor 56” (Gpr56) Cooperates with the Homeobox Gene Hoxa9 to Induce Acute Myeloid Leukemia in Mice

Interactions between leukemic cells and the bone marrow microenvironment are essential for the maintenance and progression of myeloid leukemias. G-protein coupled receptor 56 (Gpr56) is an adhesion molecule which collaborates with the extracellular matrix through interaction with collagen III binding and transglutaminase 2, and by this activating the Rho A response pathway. In the hematopoietic hierarchy, GPR56 expression is highest in stem cells and decreasing in expression with differentiation. It is a poor prognostic factor in cytogenetically normal acute myeloid leukemia (AML), and was identified as a member of a signature expressed in functionally validated human AML stem cells (Eppert et al., Nature Medicine 2011). To test its functional relevance for AML, we first analyzed the expression of GPR56 in primary patient samples (n= 74) with different genotypes from the AML dataset available in the TCGA database. Normal karyotype patients showed significantly higher expression of GPR56 (3.1 fold p<0.05), when compared to patient samples with translocations. Analysis within the normal karyotype dataset showed high GPR56 expression independent of NPM1 or FLT3-ITD mutations. Quantification of GPR56 expression by TaqMan PCR confirmed high expression both in the CD34-positive and CD34-negative leukemic subpopulations in normal karyotype AML, indicating high expression of this protein in the compartment enriched for leukemic stem cells. As GPR56 expression was high in normal karyotype AML, known to be characterized by aberrant homeobox expression, we tested the functional relevance of Gpr56 expression in a murine Hoxa9 bone marrow (BM) transplantation model. Aberrant expression of Hoxa9 in 5-FU mobilized stem and progenitor cells increased GPR56 expression 1.3 fold (p<0.05), but failed to induce short latency AML up to an observation time of over 160 days post transplant. Retrovirally engineered aberrant overexpression of Gpr56 alone also did not induce any perturbations with regard to engraftment and cell differentiation over 90 days post transplant in mice. However, overexpression of Hoxa9 together with Gpr56 induced a significant and 1.8 fold increase in colony forming cell activity (CFC) (p<0.05), followed by a 1.6 fold (p<0.05) increase in secondary CFCs when compared to HoxA9 alone. Mice transplanted with BM cells co-transduced with HoxA9 and Gpr56 rapidly developed AML after a median latency of 139 days post transplant (HoxA9 MIEG3 N=8, HoxA9 Gpr56 N=10, p<0.05). The mice displayed splenomegaly (mean 0.47g ± 0.07), and a high expression of myeloid markers on leukemic cells (Mac-1: 31%-57%; Mac-1/Gr-1: 24%-53%; Gr-1: 0.5%-16%). Cells co-transduced with Gpr56 and HoxA9 showed a decrease in Cxcr4 expression, when compared to HoxA9 alone and a 6-fold reduction in CFCs (p<0.05) when cells were introduced to an in-vitro SDF-1 gradient migration assay. However, BM cells co-transduced with Hoxa9 and Gpr56 did not show any significant drop in their homing capacity in vivo. Of note, leukemogenicity of Hoxa9 BM cells co-transduced with Meis1 was dependent on Gpr56 expression, as knockdown of Gpr56 expression via shRNA doubled the latency time in transplanted mice (median latency 55 days compared to 95 days, respectively p<0.05). Knockdown of Gpr56 induced a decrease in cells capable of attaching to the stroma in the adhesion cobble-stone area forming cell assay (CAFC) (6.5 fold reduction). However, cells able to home to the BM in vivoincreased 2.3 fold (p<0.05), suggesting that the delay in disease progression was due to a decrease in attaching to the niche.In summary, our data characterize Gpr56 as an essential partner in Hoxa9 associated myeloid leukemogenesis, underlining the importance of members of the adhesion-GPCR family of receptors in myeloid leukemogenesis. DisclosuresNo relevant conflicts of interest to declare.