141 research outputs found
On order identification of time series models and its applications
No full textMy thesis focuses on the order identification schemes of the widely-used time series model - Autoregressive Integrated Moving-Average (ARIMA) model and the applications of the order determination methods. The first part investigates the impact of dependent but uncorrelated innovations (errors) on the traditional autoregressive integrated moving average (ARIMA) model order determination schemes such as autocorrelation function (ACF), partial autocorrelation function (PACF), extended autocorrelation function (EACF) and unit-root test. We also propose a new order determination scheme to address those impacts and can be used to time series sequences with uncorrelated innovations. In the second part, a unified approach for the tentative specification of both the seasonal and nonseasonal orders of general multiplicative seasonal model is proposed. This new approach has the advantages of determining the seasonal and nonseasonal orders simultaneously and automatically. In the third part, a hierarchical model approach is presented for predicting the end-of-day stock trading volume (total daily volume). It effectively combines two sources of information: the trading volume already accumulated from the beginning of the trading day to the time of prediction, and the historical daily trading volume dynamics.Ph.D.Includes bibliographical referencesIncludes vitaby Shuhao Che
Author Name Disambiguation on Heterogeneous Information Network with Adversarial Representation Learning
No full textAuthor name ambiguity causes inadequacy and inconvenience in academic information retrieval, which raises the necessity of author name disambiguation (AND). Existing AND methods can be divided into two categories: the models focusing on content information to distinguish whether two papers are written by the same author, the models focusing on relation information to represent information as edges on the network and to quantify the similarity among papers. However, the former requires adequate labeled samples and informative negative samples, and are also ineffective in measuring the high-order connections among papers, while the latter needs complicated feature engineering or supervision to construct the network. We propose a novel generative adversarial framework to grow the two categories of models together: (i) the discriminative module distinguishes whether two papers are from the same author, and (ii) the generative module selects possibly homogeneous papers directly from the heterogeneous information network, which eliminates the complicated feature engineering. In such a way, the discriminative module guides the generative module to select homogeneous papers, and the generative module generates high-quality negative samples to train the discriminative module to make it aware of high-order connections among papers. Furthermore, a self-training strategy for the discriminative module and a random walk based generating algorithm are designed to make the training stable and efficient. Extensive experiments on two real-world AND benchmarks demonstrate that our model provides significant performance improvement over the state-of-the-art methods
Impacts of Cultivated Land Conversion on Environmental Sustainability and Grain Self-sufficiency in China
No full textUsing provincial data, the present paper examines the impact of cultivated land conversion on agriculture and the environment. It is found that the grain production center is gradually moving towards more fragile and water scarce areas, putting more pressure on the environment. Land conversion caused large losses in ecosystem service values in the 1990s, but large scale ecological restoration programs have been implemented since 2000 to compensate for such losses. The ecological restoration programs are concentrated in regions with relatively low land productivity, whereas cultivated land conversion usually takes place in areas with relatively high land productivity. Newly-cultivated land, especially that in areas marginally suit for agricultural production, is likely to have much lower productivity levels than the original cultivated land. Because the stock of potentially cultivable land is almost exhausted, China's grain self-sufficiency policy can only be maintained by preserving the available stock of arable land and increasing its productivity in a sustainable way. Copyright (c) 2008 The Author.
To explore the importance of brand to Chinese product under the international environment
No full textBrand is the sum of all tangible and intangible attributes of a product and branding is the core part of marketing which segment the market and determine the target customers of the company. Branding today is more about corporate branding which link the internal stakeholders and external stakeholders. That would allow firms to create a strong brand through integrated operation in the corporate level. Chinese manufacturing industry was criticized as “good manufacturing but no brand” and “advertising as branding”, it is valuable to investigate what is the current situation in the world second largest economy and what is the future of “made in China”. In order to analyze what is the current situation of Chinese brand and explore the feasible strategy to the industry, the author proposed a research that looks into the Chinese manufacturing industry in order to explore what are the connotation of “made in China” and where is the way for the transition economy just like China. This study mainly employs qualitative research which assisted by quantitative methods. In order to meet the research objectives, a specifically designed questionnaire accompany with several management interviews aimed at collecting primary data. The quantitative methods such as descriptive analysis, means and regression are employed to analysis the data. Through investigate the primary data; it is an urgent task for Chinese firms to use their own brand to compete in the international marketplace. To improve the quality and shorter the lead time could add more value to Chinese products (brands), and service is the main constraint for Chinese firm to create brand in the oversea market. For the further improvement, some feasible strategies are recommended from different level of strategies in order to deliver high quality product with secure time and good services. Firstly, the M&A is recommended as corporate level strategy which aims to acquire the advanced management skills, R&D assets and oversea market distribution channels. Secondly, invest on R&D and developing the own distribution channels are recommend as business level strategy. Thirdly, to implement international standard quality control system is recommended as operational level strategy
To explore the importance of brand to Chinese product under the international environment
Full text linkBrand is the sum of all tangible and intangible attributes of a product and branding is the core part of marketing which segment the market and determine the target customers of the company. Branding today is more about corporate branding which link the internal stakeholders and external stakeholders. That would allow firms to create a strong brand through integrated operation in the corporate level. Chinese manufacturing industry was criticized as “good manufacturing but no brand” and “advertising as branding”, it is valuable to investigate what is the current situation in the world second largest economy and what is the future of “made in China”. In order to analyze what is the current situation of Chinese brand and explore the feasible strategy to the industry, the author proposed a research that looks into the Chinese manufacturing industry in order to explore what are the connotation of “made in China” and where is the way for the transition economy just like China. This study mainly employs qualitative research which assisted by quantitative methods. In order to meet the research objectives, a specifically designed questionnaire accompany with several management interviews aimed at collecting primary data. The quantitative methods such as descriptive analysis, means and regression are employed to analysis the data. Through investigate the primary data; it is an urgent task for Chinese firms to use their own brand to compete in the international marketplace. To improve the quality and shorter the lead time could add more value to Chinese products (brands), and service is the main constraint for Chinese firm to create brand in the oversea market. For the further improvement, some feasible strategies are recommended from different level of strategies in order to deliver high quality product with secure time and good services. Firstly, the M&A is recommended as corporate level strategy which aims to acquire the advanced management skills, R&D assets and oversea market distribution channels. Secondly, invest on R&D and developing the own distribution channels are recommend as business level strategy. Thirdly, to implement international standard quality control system is recommended as operational level strategy
Continual Learning of Neural Machine Translation within Low Forgetting Risk Regions
Full text linkThis paper considers continual learning of large-scale pretrained neural
machine translation model without accessing the previous training data or
introducing model separation. We argue that the widely used
regularization-based methods, which perform multi-objective learning with an
auxiliary loss, suffer from the misestimate problem and cannot always achieve a
good balance between the previous and new tasks. To solve the problem, we
propose a two-stage training method based on the local features of the real
loss. We first search low forgetting risk regions, where the model can retain
the performance on the previous task as the parameters are updated, to avoid
the catastrophic forgetting problem. Then we can continually train the model
within this region only with the new training data to fit the new task.
Specifically, we propose two methods to search the low forgetting risk regions,
which are based on the curvature of loss and the impacts of the parameters on
the model output, respectively. We conduct experiments on domain adaptation and
more challenging language adaptation tasks, and the experimental results show
that our method can achieve significant improvements compared with several
strong baselines.Comment: EMNLP 2022 Main Conference Long Pape
Clinical application of optical fiber technique combined with curettage in the minimally invasive treatment of axillary osmidrosis
No full textBackground: To investigate the clinical efficacy of optical fiber technique combined with curettage method in minimally invasive treatment of axillary osmidrosis. Method: Thirty-eight patients (76 sides) who attended the author's hospital from January 2022 to December 2023 and were classified as grade 2–3 according to Park's axillary odor grading were selected as the study subjects, including 18 males and 20 females. All patients were treated with minimally invasive treatment of axillary osmidrosis using optical fiber technique combined with curettage method, and were followed up for 6–8 months after the operation, and patient data were collected, including preoperative axillary osmidrosis Park classification, postoperative efficacy, complications, recurrence rate and satisfaction survey. HE staining of the scraped tissues from six of the patients showed a large amount of sweat gland tissue. Results: The operation was successfully completed in 38 patients (76 sides). Evaluation of clinical efficacy showed that 62 patients (81.58 %) were rated as cured, 8 sides (10.53 %) were rated as significantly efficacious, 4 sides (5.26 %) were rated as effective, and 2 sides (2.63 %) were rated as ineffective. Postoperative subcutaneous fluid hematoma occurred on 2 sides, and the complication rate was 2.63 %. with a recurrence rate of 2.63 %, patient satisfaction with the surgical outcome was 94.74 %. Conclusions: Optical fiber technique combined with curettage method for minimally invasive treatment of axillary osmidrosis, with tiny incision, fast postoperative recovery, inconspicuous scar, ideal appearance, low incidence of postoperative complications and recurrence rate, and high patient satisfaction,which is worthy of clinical promotion
An NNwC MPPT-Based Energy Supply Solution for Sensor Nodes in Buildings and Its Feasibility Study
No full textSensors for data collecting are vital in the development of IoT and intelligent systems. High power consuming current and voltage monitors are indispensable in conducting maximum power point tracking (MPPT) in traditional PV energy wireless sensor nodes. This paper presents a sensor node system based on Neural Network MPPT with cloud method (NNwC) which utilizes information sharing process that is specific to sensor networks. NNwC uses a few sample sensor nodes to collect environmental parameter data such as light intensity (L) and temperature (T) to build the MPPT regression model by Neural Network. Then all other functional sensor nodes implement the model with their environmental parametervalues to conduct MPPT. As a result, the new sensor node system reduces energy consumption as well as the size and cost of the harvester. Then, this paper provides a SPICE simulation to estimate the percentage of power consumption reduced in the new sensor node system and also estimates the percentage of loss in neural network MPPT power generation compared with the perfect MPPT. Finally, the study compares the economic and environmental performance of the proposed system and the traditional ones through a case in a real building situation
Dynamically configured physics-informed neural network in topology optimization applications
Full text linkIntegration of machine learning (ML) into the topology optimization (TO)
framework is attracting increasing attention, but data acquisition in
data-driven models is prohibitive. Compared with popular ML methods, the
physics-informed neural network (PINN) can avoid generating enormous amounts of
data when solving forward problems and additionally provide better inference.
To this end, a dynamically configured PINN-based topology optimization
(DCPINN-TO) method is proposed. The DCPINN is composed of two subnetworks,
namely the backbone neural network (NN) and the coefficient NN, where the
coefficient NN has fewer trainable parameters. The designed architecture aims
to dynamically configure trainable parameters; that is, an inexpensive NN is
used to replace an expensive one at certain optimization cycles. Furthermore,
an active sampling strategy is proposed to selectively sample collocations
depending on the pseudo-densities at each optimization cycle. In this manner,
the number of collocations will decrease with the optimization process but will
hardly affect it. The Gaussian integral is used to calculate the strain energy
of elements, which yields a byproduct of decoupling the mapping of the material
at the collocations. Several examples with different resolutions validate the
feasibility of the DCPINN-TO method, and multiload and multiconstraint problems
are employed to illustrate its generalization. In addition, compared to finite
element analysis-based TO (FEA-TO), the accuracy of the displacement prediction
and optimization results indicate that the DCPINN-TO method is effective and
efficient.Comment: 31 pages, 22 figure
An efficient online successive reanalysis method for dynamic topology optimization
Full text linkIn this study, an efficient reanalysis strategy for dynamic topology
optimization is proposed. Compared with other related studies, an online
successive dynamic reanalysis method and POD-based approximate dynamic
displacement strategy are integrated. In dynamic reanalysis, the storage of the
stiffness matrix decomposition can be avoided and the reduced basis vectors
should be updated successively according to the structural status in each
iteration. Therefore, the bottleneck of combined approximation method for
large-scale dynamic topology optimization can be handled. Sequentially, the
Proper Orthogonal Decomposition (POD) is employed to obtain the approximate
dynamic displacement, in which the Proper Orthogonal Mode (POM) of the
displacement field is employed to establish the approximated equivalent static
loads of Equivalent Static Load (ESL) method. Compared with the exact
equivalent static loads at all the time intervals, the number of equivalent
static loads is significantly reduced. Finally, the 2D and 3D test results
indicate that the proposed method has remarkable speed-up effect on the premise
of small relative error, support the strength of the proposed strategy
- …
