Dimensional Sequence Research Articles

Dealing with High Dimensional Sequence Data in Manufacturing

Advanced manufacturing processes generate large number of parameters and sequences collected from multiple stages of the production pipeline. Consequently, the data dimension explodes due to the high number of features and sequence length. In such a configuration, it is extremely challenging for machine learning algorithms to perform well in limited datasets of few thousand data points. In this paper, we discuss suitable state-of-the-art techniques and propose the application and comparison of few most promising methodologies for effective feature transformation to address this challenge. The results are presented on real-world use cases from the manufacturing sector in Luxembourg.

A feature-aware SPH for isotropic unstructured mesh generation

In this paper, we present a feature-aware SPH method for the concurrent and automated isotropic unstructured mesh generation. Two additional objectives are achieved with the proposed method compared to the original SPH-based mesh generator (Fu et al., 2019). First, a feature boundary correction term is introduced to address the issue of incomplete kernel support at the boundary vicinity. The mesh generation of feature curves, feature surfaces and volumes can be handled concurrently without explicitly following a dimensional sequence. Second, a two-phase model is proposed to characterize the mesh-generation procedure by a feature-size-adaptation phase and a mesh-quality-optimization phase. By proposing a new error measurement criterion and an adaptive control system with two sets of simulation parameters, the objectives of faster feature-size adaptation and local mesh-quality improvement are merged into a consistent framework. The proposed method is validated with a set of 2D and 3D numerical tests with different complexities and scales. The results demonstrate that high-quality meshes are generated with a significant speedup of convergence.

QLens: Visual Analytics of MUlti-step Problem-solving Behaviors for Improving Question Design.

With the rapid development of online education in recent years, there has been an increasing number of learning platforms that provide students with multi-step questions to cultivate their problem-solving skills. To guarantee the high quality of such learning materials, question designers need to inspect how students' problem-solving processes unfold step by step to infer whether students' problem-solving logic matches their design intent. They also need to compare the behaviors of different groups (e.g., students from different grades) to distribute questions to students with the right level of knowledge. The availability of fine-grained interaction data, such as mouse movement trajectories from the online platforms, provides the opportunity to analyze problem-solving behaviors. However, it is still challenging to interpret, summarize, and compare the high dimensional problem-solving sequence data. In this paper, we present a visual analytics system, QLens, to help question designers inspect detailed problem-solving trajectories, compare different student groups, distill insights for design improvements. In particular, QLens models problem-solving behavior as a hybrid state transition graph and visualizes it through a novel glyph-embedded Sankey diagram, which reflects students' problem-solving logic, engagement, and encountered difficulties. We conduct three case studies and three expert interviews to demonstrate the usefulness of QLens on real-world datasets that consist of thousands of problem-solving traces.

Efficient Approaches to the Mixture Distance Problem

The ancestral mixture model, an important model building a hierarchical tree from high dimensional binary sequences, was proposed by Chen and Lindsay in 2006. As a phylogenetic tree (or evolutionary tree), a mixture tree created from ancestral mixture models, involves the inferred evolutionary relationships among various biological species. Moreover, it contains the information of time when the species mutates. The tree comparison metric, an essential issue in bioinformatics, is used to measure the similarity between trees. To our knowledge, however, the approach to the comparison between two mixture trees is still unknown. In this paper, we propose a new metric named the mixture distance metric, to measure the similarity of two mixture trees. It uniquely considers the factor of evolutionary times between trees. If we convert the mixture tree that contains the information of mutation time of each internal node into a weighted tree, the mixture distance metric is very close to the weighted path difference distance metric. Since the converted mixture tree forms a special weighted tree, we were able to design a more efficient algorithm to calculate this new metric. Therefore, we developed two algorithms to compute the mixture distance between two mixture trees. One requires O(n2) and the other requires O(nh1h2) computational time with O(n) preprocessing time, where n denotes the number of leaves in the two mixture trees, and h1 and h2 denote the heights of these two trees.

AONet: Active Offset Network for crowd flow prediction

Predicting crowd flow is of great importance to public safety and traffic management. The crowd flow is difficult to predict accurately and timely due to the uncertainty of the future positions. In this paper, we propose a novel Active Offset Network (AONet), in which ActiveGRU (Active Gate Recurrent Unit) is designed to predict the variation of pedestrians’ positions in the crowd flow. Its inner location-variant recurrent structure is implemented by utilizing convolution operation on low dimensional spatio-temporal sequences to obtain fractional offset locations. Afterwards, the sampling locations are determined by bilinear interpolation on fractional offset locations. Moreover, a probabilistic sparse strategy is introduced to reduce the links between sampling locations during supervised training. Finally, the experiments over popular benchmarks demonstrate that our method can actively characterize the future positions of pedestrians. Meanwhile, the performance of the proposed AONet is superior over state-of-art baselines with regard to both accuracy and computational savings.

Anti-Tumor TCF1+ CD8 T Cells are Functionally Diverse and Evolve During Tumorigenesis and Progression

Abstract CD8+ T cells drive protective responses against infection and cancer. In the context of chronic antigen stimulation, CD8+ T cells become progressively dysfunctional, losing the ability to proliferate, secrete cytokines, and kill target cells. While dysfunctional CD8 T cells have been characterized in infection, studies examining longitudinal responses in tumors have been limited. Here, in an autochthonous model of lung adenocarcinoma using high dimensional flow cytometry and single cell RNA sequencing, we demonstrate that tumor specific CD8 T cell heterogeneity is dynamic, revealing multiple dysfunctional-like CD8 T cell populations. Among these states, we identify TCF-1+ CD8 T cells, a population previously associated with superior functionality. Temporal analysis revealed heterogeneity within the TCF-1+ CD8+ T cell compartment - SlamF6+ CD8 T cells were predominant in early stage disease, while SlamF6- CD8 T cells appeared later and were the majority during progression. Functionally, the SlamF6+ population was proliferative and expressed more inhibitory receptors than SlamF6- cells. However, SlamF6+ CD8 T cells gradually lost the ability to divide and secrete cytokines over the course of tumorigenesis, demonstrating that they become dysfunctional over time. Collectively, our results provide new insights into longitudinal response of TCF-1+ T cells over the course of tumorigenesis and have therapeutic implications for modulating the anti-tumor T cell response.

A-081 Cognitive Reserve Factors Influence Neuropsychological Performance in Alcohol Use Disorder

Abstract Objective Cognitive reserve (CR) refers to the brain’s capacity to cope with pathology and preserve functioning. We investigated cognitive performance between individuals with alcohol use disorder (AUD) and healthy controls to examine whether CR, operationalized as education and psychosocial functioning, influences neuropsychological functioning. Method We recruited 45 AUD (DSM-V criteria) who reported drinking levels exceeding NIAAA guidelines (>14/7 drinks/week for men/women) and 30 healthy controls who did not. MANCOVAs controlling for CR were used to investigate between-group differences in neuropsychological performance, as measured by the NIH Toolbox. A series of linear regression analyses were also performed to evaluate effects of AUD and CR on neuropsychological performance. Psychosocial functioning, education, and AUD status were simultaneously entered as predictors of Flanker, Dimensional Change Card Sort, Picture Sequence, List Sort, and Processing Speed scores. Results MANCOVAs revealed a significantly slower processing speed in the AUD group compared to controls when controlling for CR (F = 4.30, p = .042). There were no significant group differences on other tests. Linear regressions showed only processing speed to be predicted by AUD (β = −.255, p = .042), while CR measures were not. Education predicted Picture Sequence (β = .245, p = .041) and Card Sort (β = .291, p = .009) performance, and psychosocial functioning predicted Flanker (β = .296, p = .021) and Card Sort (β = .316, p = .010) performance. Conclusions CR appears to contribute to higher-order cognitive functions, regardless of AUD status. Only processing speed, a domain typically susceptible to brain pathology, was significantly related to AUD. Thus, factors linked to CR may serve as important targets for future research and intervention in AUD to promote favorable cognitive outcomes.

Comparison of 3D and 2D late gadolinium enhancement magnetic resonance imaging in patients with acute and chronic myocarditis

We compared a fast, single breath-hold three dimensional LGE sequence (3D LGE) with an established two dimensional multi breath-hold sequence (2D LGE) and evaluated image quality and the amount of myocardial fibrosis in patients with acute and chronic myocarditis. 3D LGE and 2D LGE (both spatial resolution 1.5 × 1.5 mm2, slice-thickness 8 mm, field of view 350 × 350 mm2) were acquired in 25 patients with acute myocarditis (mean age 40 ± 18 years, 7 female) and 27 patients with chronic myocarditis (mean age 44 ± 22 years, 9 female) on a 1.5 T MR system. Image quality was evaluated by two independent, blinded readers using a 5-point Likert scale. Total myocardial mass, fibrotic mass and total fibrotic tissue percentage were quantified for both sequences in both groups. There was no significant difference in image quality between 3D und 2D acquisitions in patients with acute (p = 0.8) and chronic (p = 0.5) myocarditis. No significant differences between 3D and 2D acquisitions could be shown for myocardial mass (acute p = 0.2; chronic p = 0.3), fibrous tissue mass (acute p = 0.7; chronic p = 0.1) and total fibrous percentage (acute p = 0.4 and chronic p = 0.2). Inter-observer agreement was substantial to almost perfect. Acquisition time was significantly shorter for 3D LGE (24 ± 5 s) as compared to 2D LGE (350 ± 58 s, p < 0.001). In patients with acute and chronic myocarditis 3D LGE imaging shows equal diagnostic quality compared to standard 2D LGE imaging but with significantly reduced acquisition time.

A Unified Data-Adaptive Framework for High Dimensional Change Point Detection

SummaryIn recent years, change point detection for a high dimensional data sequence has become increasingly important in many scientific fields such as biology and finance. The existing literature develops a variety of methods designed for either a specified parameter (e.g. the mean or covariance) or a particular alternative pattern (sparse or dense), but not for both scenarios simultaneously. To overcome this limitation, we provide a general framework for developing tests that are suitable for a large class of parameters, and also adaptive to various alternative scenarios. In particular, by generalizing the classical cumulative sum statistic, we construct the U-statistic-based cumulative sum matrix C. Two cases corresponding to common or different change point locations across the components are considered. We then propose two types of individual test statistics by aggregating C on the basis of the adjusted Lp-norm with p ∈ {1, …, ∞}. Combining the corresponding individual tests, we construct two types of data-adaptive tests for the two cases, which are both powerful under various alternative patterns. A multiplier bootstrap method is introduced for approximating the proposed test statistics’ limiting distributions. With flexible dependence structure across co-ordinates and mild moment conditions, we show the optimality of our methods theoretically in terms of size and power by allowing the dimension d and the number of parameters q to be much larger than the sample size n. An R package called AdaptiveCpt is developed to implement our algorithms. Extensive simulation studies provide further support for our theory. An application to a comparative genomic hybridization data set also demonstrates the usefulness of our proposed methods.

Visualization of Proximal and Distal Kaplan Fibers Using 3-Dimensional Magnetic Resonance Imaging and Anatomic Dissection

Background: In current magnetic resonance imaging (MRI) of the knee, injuries to the anterolateral ligament complex (ALC) and the Kaplan fibers (KFs) are not routinely assessed. As ruptures of the KFs contribute to anterolateral rotatory instability in the anterior cruciate ligament–deficient knee, detecting these injuries on MRI may help surgeons to individualize treatment. Purpose: To visualize the KFs on 3-T MRI and to conduct a layer-by-layer dissection of the ALC. Study Design: Descriptive laboratory study. Methods: Ten fresh-frozen human cadaveric knees (mean ± SD age, 72 ± 8.5 years) without history of ligament injury were used in this study. Before layer-by-layer dissection of the ALC, MRI was performed to define the radiologic anatomy of the KFs. A coronal T1-weighted 3-dimensional turbo spin echo sequence and a transverse T2-weighted turbo spin echo sequence were obtained. Three-dimensional data sets were used for multiplanar reconstructions. Results: KFs were identified in 100% of cases on MRI and in anatomic dissection. The mean length of the proximal and distal KFs was 17.9 ± 3.6 mm and 12.4 ± 6.5 mm, respectively. On MRI, the distance from the lateral femoral epicondyle to the proximal KFs was 35.9 ± 6.9 mm and to the distal KFs, 16.6 ± 4.1 mm; in anatomic dissection, the distances were 41.4 ± 8.1 mm for proximal KFs and 28.2 ± 8.1 mm for distal KFs. The distance from the lateral joint line to the proximal KFs was 63.5 ± 7.6 mm and to the distal KFs, 45.3 ± 3.7 mm. Interobserver reliability for image analysis was excellent for all measurements. Conclusion: KFs can be consistently identified on MRI with use of 3-dimensional sequences. Subsequent anatomic dissection confirmed their close topography to the superior lateral genicular artery. For clinical implications, the integrity of the KFs should be routinely reviewed on MRI scans Clinical Relevance: As ruptures of the KFs contribute to anterolateral rotatory instability, accurate visualization of the KFs on MRI may facilitate surgical decision making for additional anterolateral procedures in the anterior cruciate ligament–deficient knee.

Stieltjes continued fractions related to the paperfolding sequence and Rudin-Shapiro sequence

We investigate two Stieltjes continued fractions given by the paperfolding sequence and the Rudin-Shapiro sequence. By explicitly describing certain subsequences of the convergents Pn(x)/Qn(x) modulo 4, we give the formal power series expansions (modulo 4) of these two continued fractions and prove that they are congruent modulo 4 to algebraic series in Z[[x]]. Therefore, the coefficient sequences of the formal power series expansions are 2-automatic. Write Qn(x)=∑i≥0an,ixi. Then (Qn(x))n≥0 defines a two-dimensional coefficient sequence (an,i)n,i≥0. We prove that the coefficient sequences (an,imod4)n≥0 introduced by both (Qn(x))n≥0 and (Pn(x))n≥0 are 2-automatic for all i≥0. Moreover, the pictures of these two dimensional coefficient sequences modulo 4 present a kind of self-similar phenomenon.

A machine learning based plasticity model using proper orthogonal decomposition

Data-driven material models have many advantages over classical numerical approaches, such as the direct utilization of experimental data and the possibility to improve performance of predictions when additional data is available. One approach to develop a data-driven material model is to use machine learning tools. These can be trained offline to fit an observed material behaviour and then be applied in online applications. However, learning and predicting history dependent material models, such as plasticity, is still challenging. In this work, a machine learning based material modelling framework is proposed for both elasticity and plasticity. The machine learning based hyperelasticity model is developed with the Feed forward Neural Network (FNN) directly whereas the machine learning based plasticity model is developed by using of a novel method called Proper Orthogonal Decomposition Feed forward Neural Network (PODFNN). In order to account for the loading history, the accumulated absolute strain is proposed to be the history variable of the plasticity model. Additionally, the strain–stress sequence data for plasticity is collected from different loading–unloading paths based on the concept of sequence for plasticity. By means of the POD, the multi-dimensional stress sequence is decoupled leading to independent one dimensional coefficient sequences. In this case, the neural network with multiple output is replaced by multiple independent neural networks each possessing a one-dimensional output, which leads to less training time and better training performance. To apply the machine learning based material model in finite element analysis, the tangent matrix is derived by the automatic symbolic differentiation tool AceGen. The effectiveness and generalization of the presented models are investigated by a series of numerical examples using both 2D and 3D finite element analysis.

Comparison of Volumetric Fat Suppressed Proton Density and Standard Routine MR Sequences in the Evaluation of Meniscal Tears

Abstract Background: The knee joint is the most common joint liable for injuries with knee MRI being the most preferred imaging modality for the assessment of knee pain. The com-monest indications are the evaluation of the meniscal lesions, and ligamentous injuries. Arthroscopy is considered the gold standard tool for the assessment of knee derangement, although knee MR exami-nation has shown significant influence upon the clinicians' management decisions. Current standard two dimensional MR protocol is time consuming with limitations of these sequences that have to be considered. Volumetric fat suppressed proton density sequence is used with the advantage of faster acquisition in comparison with standard routine sequences. The volumetric PD FS sequence can demonstrate similar diagnostic results to that of the standard three planes sequences in the assessment of internal knee derangement especially as regards meniscal injuries. Aim of Study: To assess the diagnostic accuracy of volu-metric fat suppressed proton density sequence versus standard two dimensional (2D) sequences at 1.5T MR in the evaluation of meniscal tears. Patients and Methods: The study included 81 patients presenting with suspected post traumatic meniscal injury. Patients were referred to the Kasr Al-Aini Hospital Radiology Department between February 2018 and January 2019 after full clinical examination to evaluate their knee joints using MRI. Results: In our study the results showed typical concord-ance between standard two dimensional (2D) three planes sequences and volumetric fat suppressed proton density sequence in the detection of medial and lateral meniscal tears after using arthroscopy as the gold standard. Both sequences showed sensitivity=95.92%, specificity=100%, Positive Pre-dictive Value (PPV)=100%, Negative Predictive Value (NPV) =94.12% and accuracy=97.53% for Medial Meniscal (MM) tears as well as sensitivity=75%, specificity=100% PPV=100%, NPV=94.2% and accuracy=95.06% for Lateral Meniscal (LM) tears. Conclusion: The volumetric fat suppressed proton density sequence had shorter acquisition time and delivered similar high diagnostic accuracy to the standard 2D sequences in diagnosing meniscal tears.

Differential evolution for the optimization of low-discrepancy generalized Halton sequences

Abstract Halton sequences are d–dimensional quasirandom sequences that fill the d–dimensional hyperspace in a uniform way. They can be used in a variety of applications such as multidimensional integration, uniform sampling, and, e.g., quasi–Monte Carlo simulations. Generalized Halton sequences improve the space–filling properties of original Halton sequences in higher dimensions by digit scrambling. In this work, an evolutionary optimization algorithm, the differential evolution, is used to optimize scrambling permutations of a d–dimensional generalized Halton sequence so that the discrepancy of the generated point set is minimized.

Typicality in quasispecies evolution in high dimensions.

We study quasispecies and closely related evolutionary dynamics like the replicator-mutator equation in high dimensions. In particular, we show that under certain conditions, the average fitness of almost all quasispecies of a given dimension becomes independent of mutational probabilities in high dimensional sequence spaces. This result is a consequence of concentration of measure on a high dimensional hypersphere and its extension to Lipschitz functions known as the Levy's Lemma. Our results naturally extend to other functional capabilities that can be described as Lipschitz functions and whose input parameters are the frequencies of individual constituents of the quasispecies. In order to show this, we give a generalization of Levy's Lemma and discuss possible biological consequences of our work.

Human Protein Sequence Classification using Machine Learning and Statistical Classification Techniques

In the field of computational biology, to gauge the meaningful and accurate feature for protein function predications, either the profile-based protein data or sequence-based data has been used. As we know that the prediction of enzyme class from an unknown protein is most interacted research in the current era. In this context, machine learning and statistical classification technique has been used. In this article, we have use six different machine learning and statistical classification technique such as CRT, QUEST, CHAID, C5.0, ANN and SVM for classification of 4314 number of human protein sequence data. These data are extracted form UniprotKB databank with the help of PROFEAT server. The extracted data are categorized in seven different classes. To manipulate the high dimensional protein sequence data with some missing value, the SPSS has been used for classification and estimation of the performance of classification technique. The experimental results highlight that the class C4, C5, C6 and C7 data are imbalanced that affect the overall performance of classification technique. This article provides an extensive comparative analysis of different classification technique on sequence-based protein data. The experimental analysis highlights that the SVM and C5.0 classification technique gives better result than others and can be used for protein classification and predictions.

A high throughput two-dimensional discrete cosine transform and MPEG4 motion estimation using vector coprocessor

In this work a configurable and scalable vector coprocessor for real time processing of MPEG4 motion estimation (ME) and two-dimensional DCT (2D DCT) is presented. A sequential DSP processor based on a reduced instruction set computer (RISC) processor architecture would require a frequency of 15 GHz for the real time processing of these two processes for a common intermediate format (CIF) sized sequence at 25 frames per second (fps). This frequency requirement will increase further if the image dimensions are increased. On the other hand our architecture on FPGA can achieve the real time processing rate at low frequency for CIF sized sequence and at higher frequency for full high definition (FHD) sequence for combined ME and 2D DCT. Due to configurable nature of the architecture and FPGA, this can be extended to higher dimensional image sequences. An important aspect of the architecture is that same datapath that is used for ME is also used for 2D DCT, with minor modification, leading to saving in area and time consumption. In addition the processor–coprocessor architecture has lower energy consumption and cost than the sequential processor.

Effects of acute physical activity on NIH toolbox-measured cognitive functions among children in authentic education settings

IntroductionIdentifying a dose of physical activity (PA) that can improve cognitive function in children has important implications for school-day PA recommendations. Researchers and educators have interest in this link as it relates to both health and academic performance. This study examined the dose-response relationship between PA and improvement in cognition in a sample of fifth and sixth grade students. MethodsParticipants (n = 156) from eight classes each completed two of four different cognitive assessments on an iPad, both before and after exposure to one of four randomized, 10-min PA conditions (sedentary, light, moderate, and vigorous). Conditions were standardized through use of videos to lead movement, and participants wore accelerometers to confirm fidelity to PA condition. The four cognitive assessments were selected from the NIH Toolbox Cognition Battery, and included Dimensional Change Card Sort, Flanker, Pattern Comparison, and Picture Sequence Memory tests. Hierarchical linear regression models were used to estimate the effects of condition on each test using an intention to treat analysis. ResultsFidelity to PA condition was acceptable for sedentary and light conditions, but became less precise for moderate and vigorous conditions. No significant time by condition interaction was observed for any of the cognitive assessment scores. ConclusionsResults did not substantiate a dose-response link between PA intensity and selected measures of cognitive function. More research is needed to investigate the potentially nuanced effects of short bouts of PA on cognitive functioning in children.

An Efficient Closed Maximal Pattern Sequences Mining on High Dimensional Datasets

Previous methods have presented convincing arguments that mining complete set of patterns is huge for effective usage. A compact but high quality set of patterns, such as closed patterns and maximal patterns is needed. Most of the previously maximal pattern sequences mining algorithms on high dimensional sequence, such as biological data set, work under the same support. In this paper, an efficient algorithm Closed Maximal Pattern Sequences (CMPS-Mine) for mining closed maximal patterns based on multi-support is suggested. Careful exhibitions once Beta-globin gene sequences have exhibited that CMPS-Mine expends less memory utilization and run time over Prefix Span. It generates compacted outcomes and two kinds of interesting patterns.

The comparison of direct magnetic resonance arthrography with volumetric interpolated breath-hold examination sequence and multidetector computed tomography arthrography techniques in detection of talar osteochondral lesions

ObjectiveThe aim of this study was to compare additive diagnostic values of magnetic resonance (MR) arthrography with volumetric interpolated breath-hold examination (VIBE) sequence and multidetector computed tomography (MDCT) arthrography for diagnosis and grading of talar osteochondral lesions. MethodsMDCT arthrography and MR arthrography with three dimensional VIBE sequence were performed in 27 patients. Findings of MR arthrography and MDCT arthrography images were compared with arthroscopic findings. Sensitivity, specificity, and accuracy rates were calculated for both MR arthrography and MDCT arthrography imaging findings. ResultsFor grade I osteochondral lesions; sensitivity, specificity and accuracy rates of MR arthrography were 95%, 73%, 90%, respectively; For grade I osteochondral lesions; sensitivity, specificity and accuracy rates of MDCT arthrography were 96%, 79%, 81%. For grade IV osteochondral lesions; sensitivity, specificity and accuracy rates of MDCT arthrography and MR arthrography were 100%. For grade II lesions, the sensitivity, specificity and accuracy rates of the MR arthrography were 80%, 76%, 77%, respectively; for grade III lesions, the sensitivity, specificity and accuracy rates of the MR arthrography were 78%, 68%, 75%. For grade II osteochondral lesions; the sensitivity, specificity and accuracy rates of the MDCT arthrography were 91%, 81%, 86%; for grade III osteochondral lesions; the sensitivity, specificity and accuracy rates of the MDCT arthrography were 90%, 83%, 89%; For grade II and III osteochondral lesions, MDCT arthrography had higher sensitivity, specificity and accuracy rates than MR arthrography. MDCT arthrography had higher diagnostic performance than MR arthrography for detection of grade II and III lesions (p = 0.041 and p = 0.038, respectively). ConclusionMDCT arthrography appears to be more reliable than MR arthrography with three dimensional VIBE sequence for accurate detection and grading of osteochondral lesions. Level of evidenceLevel III, Diagnostic Study.