The Experts below are selected from a list of 286344 Experts worldwide ranked by ideXlab platform
Hongbin Shen - One of the best experts on this subject based on the ideXlab platform.
-
rna protein binding motifs mining with a new hybrid deep learning based cross domain Knowledge Integration approach
BMC Bioinformatics, 2017Co-Authors: Xiaoyong Pan, Hongbin ShenAbstract:RNAs play key roles in cells through the interactions with proteins known as the RNA-binding proteins (RBP) and their binding motifs enable crucial understanding of the post-transcriptional regulation of RNAs. How the RBPs correctly recognize the target RNAs and why they bind specific positions is still far from clear. Machine learning-based algorithms are widely acKnowledged to be capable of speeding up this process. Although many automatic tools have been developed to predict the RNA-protein binding sites from the rapidly growing multi-resource data, e.g. sequence, structure, their domain specific features and formats have posed significant computational challenges. One of current difficulties is that the cross-source shared common Knowledge is at a higher abstraction level beyond the observed data, resulting in a low efficiency of direct Integration of observed data across domains. The other difficulty is how to interpret the prediction results. Existing approaches tend to terminate after outputting the potential discrete binding sites on the sequences, but how to assemble them into the meaningful binding motifs is a topic worth of further investigation. In viewing of these challenges, we propose a deep learning-based framework (iDeep) by using a novel hybrid convolutional neural network and deep belief network to predict the RBP interaction sites and motifs on RNAs. This new protocol is featured by transforming the original observed data into a high-level abstraction feature space using multiple layers of learning blocks, where the shared representations across different domains are integrated. To validate our iDeep method, we performed experiments on 31 large-scale CLIP-seq datasets, and our results show that by integrating multiple sources of data, the average AUC can be improved by 8% compared to the best single-source-based predictor; and through cross-domain Knowledge Integration at an abstraction level, it outperforms the state-of-the-art predictors by 6%. Besides the overall enhanced prediction performance, the convolutional neural network module embedded in iDeep is also able to automatically capture the interpretable binding motifs for RBPs. Large-scale experiments demonstrate that these mined binding motifs agree well with the experimentally verified results, suggesting iDeep is a promising approach in the real-world applications. The iDeep framework not only can achieve promising performance than the state-of-the-art predictors, but also easily capture interpretable binding motifs. iDeep is available at http://www.csbio.sjtu.edu.cn/bioinf/iDeep
-
rna protein binding motifs mining with a new hybrid deep learning based cross domain Knowledge Integration approach
bioRxiv, 2016Co-Authors: Xiaoyong Pan, Hongbin ShenAbstract:RNA plays important roles in cells through the interactions with proteins known as the RNA-binding proteins (RBP) and their binding motifs can provide crucial understanding of the post-transcriptional regulation of RNAs. How the RBPs correctly recognize the target RNA and why they bind specific positions is still far from clear. Artificial intelligence-based computational algorithms are widely acKnowledged to be capable of speeding up this process. Although many automatic tools have been developed to predict the RNA-protein binding sites from the rapidly growing multi-resource data, e.g. from sequence, structure data etc, their different domain specific features and formats have posed significant computational challenges. One of current most difficulties is the cross-source shared common Knowledge is usually at a higher abstraction level beyond the observed data, resulting in a low efficiency of direct Integration of observed data in different domains. The other difficulty is how to interpret the prediction results. Existing approaches tend to terminate after outputting the potential discrete binding sites on the sequence, but how to assemble them into the meaningful binding motifs is a topic worth of further investigation. In viewing of these challenges, we proposed a deep learning-based framework (iDeep) by using a novel hybrid convolutional neural network and deep belief network to predict the RBP interaction sites and motifs on RNAs. This new protocol is featured by transforming the original observed data into a high-level abstraction feature space using multiple layers of learning blocks, where the shared representations across different domains are integrated. This bottom to up abstraction strategy is also very helpful to remove the noise effects in the observed data. To validate our iDeep method, we performed experiments on 31 large-scale CLIP-seq datasets, and our results show that by integrating multiple sources of data, the average AUC can be improved by 8% compared to the best single-source-based predictor; and through cross-domain Knowledge Integration at an abstraction level, it outperforms the state-of-the-art predictors by 6%. Besides the overall enhanced prediction performance, the convolutional neural network module embedded in iDeep is also able to automatically capture the interpretable RNA sequential binding motifs for RBPs. Large-scale experiments show that these mined binding motifs agree well with the experimentally verified results, suggesting iDeep is a promising approach in the real-world applications. The iDeep is available at https://github.com/xypan1232/iDeep.
Marcia C Linn - One of the best experts on this subject based on the ideXlab platform.
-
designing virtual laboratories to foster Knowledge Integration buoyancy and density
2018Co-Authors: Jonathan M Vitale, Marcia C LinnAbstract:In this chapter, we report upon the iterative development of an online instructional unit featuring virtual laboratory activities that target the physical science concepts of density and buoyancy. We introduce a virtual laboratory activity that was designed to facilitate exploration of the relationship of mass and volume to buoyancy. We evaluate the virtual laboratory by measuring the extent to which it fosters meaningful experimentation, appropriate interpretation of evidence, and discovery of new ideas. In the first revision, we simplified the exploratory tools. This revision supported better interpretation of evidence related to a specific claim, but limiting potential for discovery of new ideas. In the second revision, we introduced an intuitive graph-based interface that allowed students to specify and rapidly test properties of virtual materials (i.e., mass and volume). This revision facilitated meaningful exploration of students’ ideas, thereby supporting both valid interpretations of evidence related to false claims and discovery of new ideas. We discuss the role that virtual laboratories can play in the design of all laboratory activities by tracking student strategies and offering opportunities to easily test new features.
-
comparing two forms of concept map critique activities to facilitate Knowledge Integration processes in evolution education
Journal of Research in Science Teaching, 2016Co-Authors: Beat Schwendimann, Marcia C LinnAbstract:Concept map activities often lack a subsequent revision step that facilitates Knowledge Integration. This study compares two collaborative critique activities using a Knowledge Integration Map (KIM), a form of concept map. Four classes of high school biology students (n ‹ 81) using an online inquiry-based learning unit on evolution were assigned to one of two conditions. Student dyads in one condition compared their concept maps against an expert map while dyads in the other condition conducted a peer-review. Analysis of the concept maps suggests that students in both conditions improved their understanding of evolution from pretest to posttest. However, the two conditions lead to different criteria: Students in the expert-map condition focused mostly on concept-focused criteria like concept classi®cation while students in the peer-review condition used more link-focused criteria like link labels and missing connections. This paper suggests that both forms of KIM critique activities can be bene®cial for constructing more coherent connections across different topics in evolution education. These results support the value of collaborative KIM critique activities and help clarify the forms of collaborative activities that are most likely to be effective to facilitate Knowledge Integration processes.
-
science learning and instruction taking advantage of technology to promote Knowledge Integration
2011Co-Authors: Marcia C Linn, Batsheva EylonAbstract:Chapter 1 - Introduction and Overview Chapter 2 - Typical Instructional Patterns Chapter 3 - Transforming Science Instruction with Technology: A Thermodynamics Case Study Chapter 4 - Particulate Structure of Matter: A Case Study Chapter 5 - Knowledge Integration Principles and Patterns Chapter 6 - Lectures and Technology Chapter 7 - Experimentation and Knowledge Integration Chapter 8 - Making Visualizations Valuable Chapter 9 - Collaboration for Knowledge Integration Chapter 10 - Professional Development for Knowledge Integration Chapter 11 - The Case for Knowledge Integration
-
science learning and instruction taking advantage of technology to promote Knowledge Integration
2011Co-Authors: Marcia C Linn, Batsheva EylonAbstract:Science Learning and Instruction describes advances in understanding the nature of science learning and their implications for the design of science instruction. The authors show how design patterns, design principles, and professional development opportunities coalesce to create and sustain effective instruction in each primary scientific domain: earth science, life science, and physical science. Calling for more in depth and less fleeting coverage of science topics in order to accomplish Knowledge Integration, the book highlights the importance of designing the instructional materials, the examples that are introduced in each scientific domain, and the professional development that accompanies these materials. It argues that unless all these efforts are made simultaneously, educators cannot hope to improve science learning outcomes. The book also addresses how many policies, including curriculum, standards, guidelines, and standardized tests, work against the goal of integrative understanding, and discusses opportunities to rethink science education policies based on research findings from instruction that emphasizes such understanding.
-
validating measurement of Knowledge Integration in science using multiple choice and explanation items
Applied Measurement in Education, 2011Co-Authors: Marcia C LinnAbstract:This study explores measurement of a construct called Knowledge Integration in science using multiple-choice and explanation items. We use construct and instructional validity evidence to examine the role multiple-choice and explanation items plays in measuring students' Knowledge Integration ability. For construct validity, we analyze item properties such as alignment, discrimination, and target range on the Knowledge Integration scale using a Rasch Partial Credit Model analysis. For instructional validity, we test the sensitivity of multiple-choice and explanation items to Knowledge Integration instruction using a cohort comparison design. Results show that (1) one third of correct multiple-choice responses are aligned with higher levels of Knowledge Integration while three quarters of incorrect multiple-choice responses are aligned with lower levels of Knowledge Integration, (2) explanation items discriminate between high and low Knowledge Integration ability students much more effectively than multipl...
Xiaoyong Pan - One of the best experts on this subject based on the ideXlab platform.
-
rna protein binding motifs mining with a new hybrid deep learning based cross domain Knowledge Integration approach
BMC Bioinformatics, 2017Co-Authors: Xiaoyong Pan, Hongbin ShenAbstract:RNAs play key roles in cells through the interactions with proteins known as the RNA-binding proteins (RBP) and their binding motifs enable crucial understanding of the post-transcriptional regulation of RNAs. How the RBPs correctly recognize the target RNAs and why they bind specific positions is still far from clear. Machine learning-based algorithms are widely acKnowledged to be capable of speeding up this process. Although many automatic tools have been developed to predict the RNA-protein binding sites from the rapidly growing multi-resource data, e.g. sequence, structure, their domain specific features and formats have posed significant computational challenges. One of current difficulties is that the cross-source shared common Knowledge is at a higher abstraction level beyond the observed data, resulting in a low efficiency of direct Integration of observed data across domains. The other difficulty is how to interpret the prediction results. Existing approaches tend to terminate after outputting the potential discrete binding sites on the sequences, but how to assemble them into the meaningful binding motifs is a topic worth of further investigation. In viewing of these challenges, we propose a deep learning-based framework (iDeep) by using a novel hybrid convolutional neural network and deep belief network to predict the RBP interaction sites and motifs on RNAs. This new protocol is featured by transforming the original observed data into a high-level abstraction feature space using multiple layers of learning blocks, where the shared representations across different domains are integrated. To validate our iDeep method, we performed experiments on 31 large-scale CLIP-seq datasets, and our results show that by integrating multiple sources of data, the average AUC can be improved by 8% compared to the best single-source-based predictor; and through cross-domain Knowledge Integration at an abstraction level, it outperforms the state-of-the-art predictors by 6%. Besides the overall enhanced prediction performance, the convolutional neural network module embedded in iDeep is also able to automatically capture the interpretable binding motifs for RBPs. Large-scale experiments demonstrate that these mined binding motifs agree well with the experimentally verified results, suggesting iDeep is a promising approach in the real-world applications. The iDeep framework not only can achieve promising performance than the state-of-the-art predictors, but also easily capture interpretable binding motifs. iDeep is available at http://www.csbio.sjtu.edu.cn/bioinf/iDeep
-
rna protein binding motifs mining with a new hybrid deep learning based cross domain Knowledge Integration approach
bioRxiv, 2016Co-Authors: Xiaoyong Pan, Hongbin ShenAbstract:RNA plays important roles in cells through the interactions with proteins known as the RNA-binding proteins (RBP) and their binding motifs can provide crucial understanding of the post-transcriptional regulation of RNAs. How the RBPs correctly recognize the target RNA and why they bind specific positions is still far from clear. Artificial intelligence-based computational algorithms are widely acKnowledged to be capable of speeding up this process. Although many automatic tools have been developed to predict the RNA-protein binding sites from the rapidly growing multi-resource data, e.g. from sequence, structure data etc, their different domain specific features and formats have posed significant computational challenges. One of current most difficulties is the cross-source shared common Knowledge is usually at a higher abstraction level beyond the observed data, resulting in a low efficiency of direct Integration of observed data in different domains. The other difficulty is how to interpret the prediction results. Existing approaches tend to terminate after outputting the potential discrete binding sites on the sequence, but how to assemble them into the meaningful binding motifs is a topic worth of further investigation. In viewing of these challenges, we proposed a deep learning-based framework (iDeep) by using a novel hybrid convolutional neural network and deep belief network to predict the RBP interaction sites and motifs on RNAs. This new protocol is featured by transforming the original observed data into a high-level abstraction feature space using multiple layers of learning blocks, where the shared representations across different domains are integrated. This bottom to up abstraction strategy is also very helpful to remove the noise effects in the observed data. To validate our iDeep method, we performed experiments on 31 large-scale CLIP-seq datasets, and our results show that by integrating multiple sources of data, the average AUC can be improved by 8% compared to the best single-source-based predictor; and through cross-domain Knowledge Integration at an abstraction level, it outperforms the state-of-the-art predictors by 6%. Besides the overall enhanced prediction performance, the convolutional neural network module embedded in iDeep is also able to automatically capture the interpretable RNA sequential binding motifs for RBPs. Large-scale experiments show that these mined binding motifs agree well with the experimentally verified results, suggesting iDeep is a promising approach in the real-world applications. The iDeep is available at https://github.com/xypan1232/iDeep.
Jasjit Singh - One of the best experts on this subject based on the ideXlab platform.
-
distributed r d cross regional Knowledge Integration and quality of innovative output
Research Policy, 2008Co-Authors: Jasjit SinghAbstract:Abstract We explore the impact of geographic dispersion of a firm's R&D activities on the quality of its innovative output. Using data on over half a million patents from 1127 firms, we find that having geographically distributed R&D per se does not improve the quality of a firm's innovations. In fact, distributed R&D appears to be negatively associated with average value of innovations. This suggests that potential gains from access to diverse ideas and expertise from different locations are, on average, offset by difficulty in achieving Integration of Knowledge across multiple locations. To investigate whether the innovating teams that do manage cross-fertilization of ideas from different locations achieve more valuable innovations, we analyze innovations for which there is evidence of such Knowledge cross-fertilization along any of the followings dimensions: Knowledge sourcing from other locations within the firm, having at least one inventor with cross-regional ties, and having at least one inventor that has recently moved from another region. Analysis along all three dimensions consistently reveals a direct positive effect cross-regional Knowledge Integration has on innovation quality, as well as a positive interaction effect of cross-regional Knowledge Integration and distributed R&D for innovation quality. More generally, our findings provide new evidence regarding the importance of cross-unit integrative mechanisms for achieving superior performance in multi-unit firms.
-
distributed r d cross regional Knowledge Integration and quality of innovative output
Social Science Research Network, 2006Co-Authors: Jasjit SinghAbstract:We explore the impact of geographic dispersion of a firm's R&D activities on the quality of its innovative output. Using data on over half a million patents from 1,127 firms, we find that having geographically distributed R&D per se does not improve the quality of a firm's innovations. In fact, distributed R&D appears to be negatively associated with average value of innovations. This suggests that potential gains from access to diverse ideas and expertise from different locations probably gets offset by difficulty in achieving Integration of Knowledge across multiple locations. To investigate whether the innovating teams that do manage cross-fertilization of ideas from different locations achieve more valuable innovations, we analyze innovations for which there is evidence of such Knowledge cross-fertilization along one of the followings dimensions: Knowledge sourcing from remote R&D units, having at least one inventor with cross-regional ties, and having at least one inventor that has recently moved from another region. Analysis along these three dimensions consistently reveals a positive relationship between cross-regional Knowledge Integration and quality of resulting innovations. More generally, our findings provide new evidence regarding the importance of cross-unit integrative mechanisms for achieving superior performance in multi-unit firms.
Sue Newell - One of the best experts on this subject based on the ideXlab platform.
-
social capital and Knowledge Integration in an erp project team the importance of bridging and bonding
British Journal of Management, 2004Co-Authors: Sue Newell, C Tansley, Jimmy C HuangAbstract:A project team, set up to design and implement a large-scope IT system, is essentially tasked with integrating distributed Knowledge. This suggests that the social capital of members will be organizationally important. However, we suggest that in understanding the relationship between social capital and Knowledge Integration within a project team, it is necessary to distinguish between two forms of social capital – external bridging social capital and internal bonding social capital. We argue that for the effective mobilization of ‘weak’ social capital bridges for collective purposes, there is first a need to create ‘strong’ social capital bonds within the project team so that it becomes a cohesive social unit that will be able to effectively integrate Knowledge that is acquired through members' bridging activity.
-
Knowledge Integration processes and dynamics within the context of cross functional projects
International Journal of Project Management, 2003Co-Authors: Jimmy C Huang, Sue NewellAbstract:This paper examines the dynamics of Knowledge Integration in the context of cross-functional project implementation within four large organizations. Specifically, the research focuses on exploring and conceptualizing the efficiency, scope and flexibility of Knowledge Integration [Organization Science, 7(4) (1996), 375] of which limited empirical evidence has been offered. Through the comparative study, the findings suggest that Knowledge Integration in the context of cross-functional project implementation is in essence a process of engaging organizational members through the promotion of project benefits and the management of social networks. Also, our findings reveal that an organization's embedded practices, past Integration experience and social capital plays a key role in shaping the level of coordination that in turn influences the efficiency and scope of Integration. In particular, the development and nurturing of social capital within and beyond the project team is crucial, as is the promotion of project awareness through the creation of common Knowledge.
-
Knowledge Integration as a key problem in an erp implementation
International Conference on Information Systems, 2001Co-Authors: Sue Newell, Jimmy C Huang, Alvin Wan Kok CheungAbstract:While previous studies have focused mainly on the potential benefits and critical success factors associated with ERP implementation, very few have explored the important issues of impediments encountered, especially from a Knowledge Integration perspective. We have adopted a Knowledge Integration view that focuses not on the distribution and adoption of particular technological artifacts (ERP systems), but on the Knowledge Integration processes involved in implementation. The focus of this case study is to understand the nature, structure and process of Knowledge Integration that occurs during ERP implementation. The paper has identified the Integration of Knowledge as a key problem in ERP implementation. We discovered four reasons: (1) Knowledge is embedded in complex organizational processes; (2) Knowledge is embedded in legacy systems; (3) Knowledge is embedded in externally based processes; and (4) Knowledge is embedded in the ERP system. Based on our analysis, we further suggest that to overcome these impediments to Knowledge Integration requires the development of interpersonal relations (one-to-one based) and community relations (group-based).
