Conventional Comparison Methods Research Articles

Zeroth- and first-order difference discrimination for unsupervised domain adaptation

Unsupervised domain adaptation transfers empirical knowledge from a label-rich source domain to a fully unlabeled target domain with a different distribution. A core idea of many existing approaches is to reduce the distribution divergence between domains. However, they focused only on part of the discrimination, which can be categorized into optimizing the following four objectives: reducing the intraclass distance between domains, enlarging the interclass distances between domains, reducing the intraclass distances within domains, and enlarging the interclass distances within domains. Moreover, because few methods consider multiple types of objectives, the consistency of data representations produced by different types of objectives has not yet been studied. In this paper, to address the above issues, we propose a zeroth- and first-order difference discrimination (ZFOD) approach for unsupervised domain adaptation. It first optimizes the above four objectives simultaneously. To improve the discrimination consistency of the data across the two domains, we propose a first-order difference constraint to align the interclass differences across domains. Because the proposed method needs pseudolabels for the target domain, we adopt a recent pseudolabel generation method to alleviate the negative impact of imprecise pseudolabels. We conducted an extensive comparison with nine representative conventional methods and seven remarkable deep learning-based methods on four benchmark datasets. Experimental results demonstrate that the proposed method, as a conventional approach, not only significantly outperforms the nine conventional comparison methods but is also competitive with the seven deep learning-based comparison methods. In particular, our method achieves an accuracy of 93.4% on the Office+Caltech10 dataset, which outperforms the other comparison methods. An ablation study further demonstrates the effectiveness of the proposed constraint in aligning the objectives.

Compton sequence estimation based on the deep learning method

Determining the sequence of Compton scattering and photoelectric absorption events for a Compton camera system through timing information is difficult due to the finite timing resolution of radiation detectors. The conventional method compares the energies of two sequential events and determines the order of these events. The deep learning method can estimate the sequence of Compton scattering followed by the photoelectric effect better than the conventional method because it determines the sequence based on both energy and positional information of the radiation interaction. The initial information of the deep learning models is the position and energy information, and the input data are then processed in the nodes of the hidden layers. In this study, the performance of deep learning models for Compton sequence estimation and the effect of position information on these methods were investigated. The accuracies of the deep learning method and the conventional comparison method were compared. The weights connecting each node were analyzed to evaluate the effects of position and energy information to determine the Compton sequence.

A Study on the Quality Improvement of Language Teaching in the IoT Environment

IoT networks and Teaching System have attracted much attention from researchers of industry and academia. As a typical application of IoT tworks and Teaching System, the system studied in this paper is a three-way platform for IoT technology, language teaching methods, and language teaching quality is incorporated into the framework of language teaching system. Based on the current stage of language teaching methods in China and its digital needs, the role of IoT-based language teaching models on language teaching quality improvement is investigated by establishing an IoT language teaching model and using conventional comparison methods to compare conventional language teaching quality improvement models with language teaching models based on IoT platforms to consider the impact of IoT technology on language teaching quality improvement. The results show that in the educational context where digitalization and intelligence are the trends, the intelligent language teaching model integrating IoT technology is much better than the language teaching model under the conventional model in terms of teaching quality, and through IoT technology, more language enthusiasts can become language learners and improve their learning ability. The work in this paper provides an important application guidance to IoT networks and Teaching System.

Developing new automated alternation flicker using optic disc photography for the detection of glaucoma progression.

PurposeTo evaluate a progression-detecting algorithm for a new automated matched alternation flicker (AMAF) in glaucoma patients.MethodsOpen-angle glaucoma patients with a baseline mean deviation of visual field (VF) test>-6 dB were included in this longitudinal and retrospective study. Functional progression was detected by two VF progression criteria and structural progression by both AMAF and conventional comparison methods using optic disc and retinal nerve fiber layer (RNFL) photography. Progression-detecting performances of AMAF and the conventional method were evaluated by an agreement between functional and structural progression criteria. RNFL thickness changes measured by optical coherence tomography (OCT) were compared between progressing and stable eyes determined by each method.ResultsAmong 103 eyes, 47 (45.6%), 21 (20.4%), and 32 (31.1%) eyes were evaluated as glaucoma progression using AMAF, the conventional method, and guided progression analysis (GPA) of the VF test, respectively. The AMAF showed better agreement than the conventional method, using GPA of the VF test (κ=0.337; P<0.001 and κ=0.124; P=0.191, respectively). The rates of RNFL thickness decay using OCT were significantly different between the progressing and stable eyes when progression was determined by AMAF (-3.49±2.86 μm per year vs -1.83±3.22 μm per year; P=0.007) but not by the conventional method (-3.24±2.42 μm per year vs -2.42±3.33 μm per year; P=0.290).ConclusionsThe AMAF was better than the conventional comparison method in discriminating structural changes during glaucoma progression, and showed a moderate agreement with functional progression criteria.

Differential gene expression profile of first-generation and second-generation rapamycin-resistant allogeneic T cells

We completed a phase II clinical trial evaluating rapamycin-resistant allogeneic T cells (T-rapa) and now have evaluated a T-rapa product manufactured in 6 days (T-rapa(6)) rather than 12 days (T-Rapa(12)). Using gene expression microarrays, we addressed our hypothesis that the two products would express a similar phenotype. The products had similar phenotypes using conventional comparison methods of cytokine secretion and surface markers. Unsupervised analysis of 34,340 genes revealed that T-rapa(6) and T-rapa(12) products clustered together, distinct from culture input CD4(+) T cells. Statistical analysis of T-rapa(6) products revealed differential expression of 19.3% of genes (n = 6641) compared with input CD4(+) cells; similarly, 17.8% of genes (n = 6147) were differentially expressed between T-rapa(12) products and input CD4(+) cells. Compared with input CD4(+) cells, T-rapa(6) and T-rapa(12) products were similar in terms of up-regulation of major gene families (cell cycle, stress response, glucose catabolism, DNA metabolism) and down-regulation (inflammatory response, immune response, apoptosis, transcriptional regulation). However, when directly compared, T-rapa(6) and T-rapa(12) products showed differential expression of 5.8% of genes (n = 1994; T-rapa(6) vs. T-rapa(12)). Second-generation T-rapa(6) cells possess a similar yet distinct gene expression profile relative to first-generation T-rapa(12) cells and may mediate differential effects after adoptive transfer.

Defect detection on images using multiple reference images: solving a binary labeling problem using graph-cuts algorithm

We present a new formulation to solve a defect detection problem on images using multiple reference images. The reference images are defect-free images obtained from the same position of other products. The defect detection problem is reformulated as a binary labeling problem, where each pixel is labeled with one if it contains a defect and with zero otherwise. The formulation of the energy function used for the labeling problem is defined. Then, the graph-cuts algorithm is used to obtain the optimal label set minimizing the energy function that becomes the defect detection result. The presented approaches are robust to noises taken from several sources, including image-taking, transmission process, environmental lighting, and pattern variation. It does not suffer from the alignment problem for the conventional comparison methods using references. These approaches are illustrated with real data sets, semiconductor wafer images collected by scanning electron microscope equipment, and compared to other defect detection approach.

High‐frequency characterization of packaging network in TO‐CAN photodiode modules

AbstractA simple method for analyzing the effects of TO packaging network on the high‐frequency response of photodiode modules is presented. This method is established based on the relations of the scattering parameters of the packaging network, photodiode chip, and module. It is shown that the results obtained by this method agree well with those obtained by the conventional comparison method. The proposed method is much more convenient since only the electrical domain measurements are required. © 2008 Wiley Periodicals, Inc. Microwave Opt Technol Lett 50: 1219–1223, 2008; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/mop.23356

Determination of trace halogens in rock samples by radiochemical neutron activation analysis coupled with the k0-standardization method

Radiochemical neutron activation analysis coupled with the k 0-standardization method ( k 0-RNAA method) was applied to silicate rock samples for the simultaneous determination of trace halogens (Cl, Br and I). Analytical results obtained by the k 0-RNAA method for geological standard rocks and meteorite samples agreed with those determined by the conventional comparison method conducted in the same set of experiments, suggesting that the k 0-RNAA method is as reliable as the conventional method. Our data for these samples are in good agreement with their literature values except for rare cases. Detection limits calculated under the present experimental condition are sufficiently low for Cl and Br but not for I for typical geologic and meteoritic samples. The k 0-RNAA method coupled with longer neutron-irradiation is expected to yield satisfactorily low detection limits for halogens including I in these samples.

Mixed monotone decomposition of dynamical systems with application

The notions of mixed monotone decomposition of dynamical systems are introduced. The fundamental idea is to make an elaborate use of the natural growth and decay mechanism inherent in the structure of a dynamical systems to decompose its dynamics into increase and decrease parts, and thereby to constitute an augmented dynamical system as the secalled “two-sided comparison system” of the original one. The corresponding two-sided comparison theorems show that the solution of the comparison system gives lower and upper bounds of that of the original system. Therefore, the properties of a dynamical system can be obtained through the study of its two-sided comparison system. Compared with the conventional comparison method in literature, the mixed monotone decomposition method developed herein takes in a natural way structural properties of dynamical systems into account instead of requiring strict conditions of (quasi-)monotonicity on them, and could yields refined, usually nonsymmetrical, state estimates, and thus is more suitable for systems with nonsymmetrical state constraints. As an application of the method, a sufficient condition is established for the global asymptotic stability of the trivial solution of a class of continuous-time systems with nonsymmetrical state saturation. The condition is given in terms of coefficients and state saturation levels of such systems, and contains as a special case a recent result on systems with symmetric state saturation in literature.

Continuous similarity measure between nonoverlapping X‐ray powder diagrams of different crystal modifications

AbstractThe problem of quantifying similarity between crystal structures is transformed into the problem of comparing the associated X‐ray powder diagrams. A smooth similarity measure between two powder diagrams, termed a “fold,” is defined. In contrast to conventional comparison methods, the introduced method is still applicable when the peaks of the spectra to be compared have no overlap. The main areas of application of the method are the construction of a molecular crystal structure when only the experimental powder diagram is available and the analysis of possible crystal packings predicted on the basis of molecular information only. A suitable empirical parameterization of the fold has been derived from a large set of experimental and force‐field‐generated crystals. The analysis of the outcome of an ab initio packing of a flexible molecule is given as an example. The algorithmic details of the method are given as a FORTRAN 77 code. © John Wiley &amp; Sons, Inc.

Effective recombination coefficient in theEregion

On the assumption that the ion production rate at E-layer maximum varies diurnally with the cosine of the solar zenith angle, a relative value of an ‘instantaneous’ recombination coefficient can be derived which changes only slightly with the time of day. These changes are attributed to the effect of vertical ion drift due to the Sq current system. Consequently, the absolute value of the true recombination coefficient can be derived with the conventional comparison method at the (appropriately redefined) focus of the Sq current system. An estimate of the drift perturbations of the coefficient can be obtained by using the proportionality between the drift gradient and the magnetic north component (X′) of the Sq variations in the horizontal magnetic field on the ground. The relatively small diurnal variations found for the value of the instantaneous recombination coefficient in contrast with the large diurnal variation obtained by the ‘comparison’ methods indicate that it is possible to think of a single daytime value of the coefficient for this region. The implications of this single value (∼1×10−8 cm3/sec) in terms of ionic chemistry are also examined.