An official website of the United States government
Accurate modeling of student knowledge is essential for large-scale online learning systems that are increasingly used for student training. Knowledge tracing aims to model student knowledge state given the student's sequence of learning activities. Modern Knowledge tracing (KT) is usually formulated as a supervised sequence learning problem to predict students' future practice performance according to their past observed practice scores by summarizing student knowledge state as a set of evolving hidden variables. Because of this formulation, many current KT solutions are not fit for modeling student learning from non-assessed learning activities with no explicit feedback or score observation (e.g., watching video lectures that are not graded). Additionally, these models cannot explicitly represent the dynamics of knowledge transfer among different learning activities, particularly between the assessed (e.g., quizzes) and non-assessed (e.g., video lectures) learning activities. In this paper, we propose Transition-Aware Multi-activity Knowledge Tracing (TAMKOT), which models knowledge transfer between learning materials, in addition to student knowledge, when students transition between and within assessed and non-assessed learning materials. TAMKOT is formulated as a deep recurrent multi-activity learning model that explicitly learns knowledge transfer by activating and learning a set of knowledge transfer matrices, one for each transition type between student activities. Accordingly, our model allows for representing each material type in a different yet transferrable latent space while maintaining student knowledge in a shared space. We evaluate our model on three real-world publicly available datasets and demonstrate TAMKOT's capability in predicting student performance and modeling knowledge transfer.
more »
« less
Efficient Distortion Prediction of Additively Manufactured Parts Using Bayesian Model Transfer Between Material Systems
Abstract Distortion in laser-based additive manufacturing (LBAM) is a critical issue that adversely affects the geometric integrity of additively manufactured parts and generally exhibits a complicated dependence on the underlying material. The differences in properties between distinct materials prevent the immediate application of a distortion model learned for one material to another, which introduces the challenge in LBAM of learning a distortion model for a new material system given past experiments. Current methods for investigating the distortion of different material systems typically involve finite element analysis or a large number of experiments in an empirical study. However, these methods do not learn from previous experiments and can incur significant costs in terms of computation, time, or resources. We propose a Bayesian model transfer methodology that is both physics-based and data-driven to leverage past experiments on previously studied material systems for more efficient distortion modeling of new systems. This method transfers distortion models across distinct materials based on the statistical effect equivalence framework by formulating the differences between two materials as a lurking variable. Our method reduces the experimentation and effort needed for specifying distortion models for new material systems. We validate our methodology in a case study of distortion model transfer from Ti–6Al–4V disks to 316L stainless steel disks. This case study is the first instance of model transfer between material systems and illustrates the ability of the Bayesian model transfer methodology to address the issue of comprehensive distortion modeling across varying material systems in LBAM.
more »
« less
- Award ID(s):
- 1635966
- PAR ID:
- 10173698
- Date Published:
- Journal Name:
- Journal of Manufacturing Science and Engineering
- Volume:
- 142
- Issue:
- 5
- ISSN:
- 1087-1357
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Wire arc additive manufacturing is a promising additive manufacturing process because of its high deposition rate, and material diversity. However, the low quality of melted parts is a critical issue, owing to the difficulty in establishing design rules for process–structure–property–performance. Previous studies have resolved this challenge by deriving anomaly detection models for quality monitoring and have largely relied on machine learning by training melt pool image data. Acquiring sufficient data is a key to obtaining reliable models in machine learning; however, an issue arises from concerning the cost intensiveness in high-cost materials. We propose a material-adaptive anomaly detection method to detect balling defects in a target material using property-concatenated transfer learning. First, transfer learing is applied to derive convolutional neural network (CNN)-based models from a source material and transfer them to a target material, wherein data are insufficient and machine learning rarely achieves high performance. Second, material properties are concatenated on transfer learning as additional features onto image features, contrary to typical transfer learning where CNNs only extract image features. We perform experiments in a gas tungsten arc welding system with low-carbon steel (LCS), stainless steel (STS), and inconel (INC) materials. Our models achieve best classification accuracies of 82.95%, 89.47%, and 84.22% when transferring from LCS to STS, LCS to INC, and STS to INC, respectively, compared with 78.03%, 86.37%, and 73.63% obtained using typical transfer learning. The proposed method can effectively resolve the data scarcity by model transfer from sufficient datasets in low-cost materials to rare datasets in high-cost materials. Moreover, it outperforms typical transfer learning because material properties are learned as manufacturing-knowledge features, accounting for melting and hardening characteristics of materials.more » « less
-
Abstract We present Markov Chain Monte Carlo radiative transfer modeling of a joint ALMA 345 GHz and spectral energy distribution data set for a sample of 97 protostellar disks from the VLA and ALMA Nascent Disk and Multiplicity Survey of Orion Protostars. From this modeling, we derive disk and envelope properties for each protostar, allowing us to examine the bulk properties of a population of young protostars. We find that disks are small, with a median dust radius of 29.4 − 2.7 + 4.1 au and a median dust mass of 5.8 − 2.7 + 4.6 M ⊕ . We find no statistically significant difference between most properties of Class 0, Class I, and flat-spectrum sources with the exception of envelope dust mass and inclination. The distinction between inclination is an indication that the Class 0/I/flat-spectrum system may be difficult to tie uniquely to the evolutionary state of protostars. When comparing with Class II disk dust masses in Taurus from similar radiative transfer modeling, we further find that the trend of disk dust mass decreasing from Class 0 to Class II disks is no longer present, though it remains unclear whether such a comparison is fair owing to differences in star-forming region and modeling techniques. Moreover, the disks we model are broadly gravitationally stable. Finally, we compare disk masses and radii with simulations of disk formation and find that magnetohydrodynamical effects may be important for reproducing the observed properties of disks.more » « less
-
Many materials systems comprise complex structures where multiple materials are integrated to achieve a desired performance. Often in these systems, it is a combination of both the materials and their structure that dictate performance. Here the authors layout an integrated computational–statistical–experimental methodology for hierarchical materials systems that takes a holistic design approach to both the material and structure. The authors used computational modeling of the physical system combined with statistical design of experiments to explore an activated carbon adsorption bed. The large parameter space makes experimental optimization impractical. Instead, a computational–statistical approach is coupled with physical experiments to validate the optimization results.more » « less
-
null (Ed.)Model-Based Systems Engineering (MBSE) is the formalized application of modeling to support various system evolving stages starting from the conceptual design phase to all the life cycle phases that follow. To facilitate in an efficient system behavior design process, in this paper, a Design Structure Matrix (DSM) based approach is developed and illustrated for determining the operational sequence of activities relevant to requirements of the system and in identifying the concurrent activities as well. A triangularization algorithm method is extended especially for application on activity diagrams to determine knowledge activities in an interaction graph to identify groups of activities and arrange them concurrently. The findings through the DSM based approach are validated by a system engineering expert and are implemented to construct the MBSE activity diagram to facilitate an enhanced behavior design of the system. This paper illustrates the use of Design Structure Matrix to facilitate modeling interdependencies between activities and the approach to aggregate the resulted sequential and concurrent activities with the activity diagram, applied to a case study of Execute Hohmann Transfer based on DellSat-77 Satellite System. In addition, the potentials benefits of using a Design Structure Matrix methodology for assisting Model Based Systems Engineering activities for enhanced systems behavior design is portrayed.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1115/1.4046408
