Abstract ：

Knowledge graph (KG) is evolving rapidly and play an important role in many applications. Recently, the few-shot knowledge graph completion (FKGC) task, which involves predicting missing information based on a limited number of known facts, has garnered growing interest from both practitioners and researchers. However, these methods often fail to fully utilize neighboring information and overlook the semantic distance between similar entities. To address these problems, we introduce a multi-hop neighbor aggregator based on CNN which designs to make comprehensive use of neighbor information. Additionally, we employ contrastive learning to reduce the semantic distance between similar entities. Compared to the leading baseline GANA, our model shows an improvement of 0.4% and 0.4% on NELL in terms of Hits@1 and Hits@5, respectively, and 5%, 1.4%, and 0.4% on Wiki in terms of Hits@1, Hits@5, and Hits@10. Extensive experiments demonstrate that our method performs exceptionally well on two public datasets. © 2024 SPIE.

Keyword ：

Knowledge graph Federated learning Zero-shot learning Adversarial machine learning Contrastive Learning

Cite：

Copy from the list or Export to your reference management。

Abstract ：

Binary code representation learning has shown significant performance in binary analysis tasks. However, existing solutions often have poor transferability, particularly in few-shot and zero shot scenarios where few or no training samples are available for the tasks. To address this problem, we present CLAP (Contrastive Language-Assembly Pre-training), which employs natural language supervision to learn better representations of binary code (i.e., assembly code) and get better transferability. At the core, our approach boosts superior transfer learning capabilities by effectively aligning binary code with their semantic explanations (in natural language), resulting in a model able to generate better embeddings for binary code. To enable this alignment training, we propose an efficient data generator that can automatically generate a large and diverse dataset comprising binary code and corresponding natural language explanations. We have generated 195 million pairs of binary code and explanations and trained a prototype of CLAP. The evaluations of CLAP across various downstream tasks in binary analysis all demonstrate exceptional performance. Notably, without any task-specific training, CLAP is often competitive with a fully supervised baseline, showing excellent transferability.

Keyword ：

Representation Learning Binary Analysis Deep Learning

Cite：

Copy from the list or Export to your reference management。

Abstract ：

China joined the Paris Agreement, and the global 2 degrees C and 1.5 degrees C warming targets will be supported by China. In order to achieve these targets, China's CO2 emissions need to be cut deeply by 2050. The present paper presents studies from the integrated policy assessment model for China (IPAC) team about the impact on China's economic development of deep cuts in greenhouse gas (GHG) emissions, in order to realize the Paris climate change targets. With the requirement of deep cuts in GHG emissions in China, China's economic development will also be impacted in moving toward a low-carbon or zero-carbon emission-based economy by 2050. This means the Chinese economy needs a strong transition over the next three decades, a relatively short time. All sectors in the economy need to seek ways to reduce GHG emissions, and this could change activities, industry processes and technologies in order to make the deep cuts in GHG emissions happen. This is the meaning of the economic transition toward to a low-carbon economy. The findings of the present paper include: a significant transition in the energy supply sector; a high rate of electrification in all end-use sectors; and a technology transition in the transport sector. Transitions will also occur in the traditional industrial sectors, including steel making, cement manufacture, and the chemical sector. The availability of low-cost renewable energy could change the allocation of industries, which could potentially have a strong impact on regional economic development. Deep cuts in CO2 emissions in China need not be a burden for economic development, as the IPAC results show there will be a more than 1.5% increase of gross domestic product by 2050 in the deep cut scenario compared with the baseline scenario.

Keyword ：

Paris Agreement scenario China energy transition economy transition CO2 emission reduction

Cite：

Copy from the list or Export to your reference management。

Abstract ：

Ammonia (NH3) discharged from agricultural activities to the atmosphere plays a crucial role in the formation of secondary inorganic aerosols. This study analyzed the temporal-spatial development of agricultural NH3 emissions from 2000 to 2018 in the Beijing-Tianjin-Hebei (BTH) region and assessed the effects of reducing PM2.5 by removing agricultural NH3 using an air quality model. The results showed that the interannual agricultural NH3 emissions in the BTH region exhibited a stairs trend from 2000 to 2018, with an average of 971.63 Gg. In particular, agricultural NH3 emissions in the BTH region reached a maximum in summer when the temperature was high and were more concentrated in the southern plains compared to the northern areas. Under the reduction scenario (RS), the agricultural NH3 emissions in the BTH region in 2015, 2016, 2017, and 2018 were reduced by 2.95%, 4.10%, 18.75%, and 10.21%, resulting in a reduction of 0.5%, 0.5%, 2.5%, and 1.2% of annual mean PM2.5 concentration, respectively, compared with the baseline scenario (BS). Furthermore, agricultural NH3 emissions contributed 12.6, 12.1, 11.9, and 11.3 mu g m(-3) to PM2.5 concentrations in 2015, 2016, 2017, and 2018 under the zero-emission scenario (ZS), respectively. However, the contribution rates exhibited a slightly increasing trend from 20.5% in 2015 to 24.6% in 2018. These findings could provide a new understanding of agricultural NH3 emission trends and their impacts on PM2.5 concentration based on actual NH3 mitigation ratios in recent years, thereby guiding the formulation of future control strategies.

Keyword ：

PM2.5 Ammonia Reduction effects Emission trends

Cite：

Copy from the list or Export to your reference management。

Abstract ：

In this paper, we propose to leverage the emerging deep learning techniques for spatiotemporal modeling and prediction in cellular networks, based on big system data. First, we perform a preliminary analysis for a big dataset from China Mobile, and use traffic load as an example to show non-zero temporal autocorrelation and non-zero spatial correlation among neighboring Base Stations (BSs), which motivate us to discover both temporal and spatial dependencies in our study. Then we present a hybrid deep learning model for spatiotemporal prediction, which includes a novel autoencoder-based deep model for spatial modeling and Long Short-Term Memory units (LSTMs) for temporal modeling. The autoencoder-based model consists of a Global Stacked AutoEncoder (GSAE) and multiple Local SAEs (LSAEs), which can offer good representations for input data, reduced model size, and support for parallel and application-aware training. Moreover, we present a new algorithm for training the proposed spatial model. We conducted extensive experiments to evaluate the performance of the proposed model using the China Mobile dataset. The results show that the proposed deep model significantly improves prediction accuracy compared to two commonly used baseline methods, ARIMA and SVR. We also present some results to justify effectiveness of the autoencoder-based spatial model.

Keyword ：

Autoencoder Big Data Cellular Network Deep Learning Recurrent Neural Network Spatiotemporal Modeling

Cite：

Copy from the list or Export to your reference management。

Abstract ：

Classical least squares (CLS) regression is a popular multivariate statistical method used frequently for quantitative analysis using Fourier transform infrared (FT-IR) spectrometry. Classical least squares provides the best unbiased estimator for uncorrelated residual errors with zero mean and equal variance. However, the noise in FT-IR spectra, which accounts for a large portion of the residual errors, is heteroscedastic. Thus, if this noise with zero mean dominates in the residual errors, the weighted least squares (WLS) regression method described in this paper is a better estimator than CLS. However, if bias errors, such as the residual baseline error, are significant, WLS may perform worse than CLS. In this paper, we compare the effect of noise and bias error in using CLS and WLS in quantitative analysis. Results indicated that for wavenumbers with low absorbance, the bias error significantly affected the error, such that the performance of CLS is better than that of WLS. However, for wavenumbers with high absorbance, the noise significantly affected the error, and WLS proves to be better than CLS. Thus, we propose a selective weighted least squares (SWLS) regression that processes data with different wavenumbers using either CLS or WLS based on a selection criterion, i.e., lower or higher than an absorbance threshold. The effects of various factors on the optimal threshold value (OTV) for SWLS have been studied through numerical simulations. These studies reported that: (1) the concentration and the analyte type had minimal effect on OTV; and (2) the major factor that influences OTV is the ratio between the bias error and the standard deviation of the noise. The last part of this paper is dedicated to quantitative analysis of methane gas spectra, and methane/toluene mixtures gas spectra as measured using FT-IR spectrometry and CLS, WLS, and SWLS. The standard error of prediction (SEP), bias of prediction (bias), and the residual sum of squares of the errors (RSS) from the three quantitative analyses were compared. In methane gas analysis, SWLS yielded the lowest SEP and RSS among the three methods. In methane/toluene mixture gas analysis, a modification of the SWLS has been presented to tackle the bias error from other components. The SWLS without modification presents the lowest SEP in all cases but not bias and RSS. The modification of SWLS reduced the bias, which showed a lower RSS than CLS, especially for small components.

Keyword ：

baseline error classical least squares CLS Fourier transform infrared FT-IR noise selective weighted least squares SWLS weighted least squares WLS

Cite：

Copy from the list or Export to your reference management。

Abstract ：

In this paper, we propose to leverage the emerging deep learning techniques for spatiotemporal modeling and prediction in cellular networks, based on big system data. First, we perform a preliminary analysis for a big dataset from China Mobile, and use traffic load as an example to show non-zero temporal autocorrelation and non-zero spatial correlation among neighboring Base Stations (BSs), which motivate us to discover both temporal and spatial dependencies in our study. Then we present a hybrid deep learning model for spatiotemporal prediction, which includes a novel autoencoder-based deep model for spatial modeling and Long Short-Term Memory units (LSTMs) for temporal modeling. The autoencoder-based model consists of a Global Stacked AutoEncoder (GSAE) and multiple Local SAEs (LSAEs), which can offer good representations for input data, reduced model size, and support for parallel and application-aware training. Moreover, we present a new algorithm for training the proposed spatial model. We conducted extensive experiments to evaluate the performance of the proposed model using the China Mobile dataset. The results show that the proposed deep model significantly improves prediction accuracy compared to two commonly used baseline methods, ARIMA and SVR. We also present some results to justify effectiveness of the autoencoder-based spatial model. © 2017 IEEE.

Keyword ：

Deep neural networks Wireless networks Big data Recurrent neural networks Cellular neural networks Forecasting Learning systems Deep learning Mobile telecommunication systems

Cite：

Copy from the list or Export to your reference management。

Abstract ：

On bandwidth-limited visible light communication (VLC) transmission systems, direct current (DC) component loss, DC-unbalance of code, and severe high-frequency attenuation cause baseline wander (BLW) and data-dependent jitter (DDJ) phenomena, which deteriorate signal quality and result in a higher bit error rate (BER). We present a scheme based on hybrid run length limited codes and pre-emphasis techniques to decrease the inter-symbol interference caused by BLW and DDJ phenomena. We experimentally demonstrate, utilizing 1-binary-digit-into-2-binary-digits (1B2B) codes and postcursor pre-emphasis techniques, that the impacts of BLW and DDJ on on-off keying nonreturn-to-zero VLC systems are alleviated and a 130 Mb/s data transmission rate with a BER performance of <10(-4) can be achieved. (C) 2016 Society of Photo-Optical Instrumentation Engineers (SPIE).

Keyword ：

visible light communication pre-emphasis run length limited code data-dependent-jitter baseline wander direct current-balance

Cite：

Copy from the list or Export to your reference management。

Abstract ：

As the result of climate change and deteriorating global environmental quality, nations are under pressure to reduce their emissions of greenhouse gases per unit of GDP. China has announced that it is aiming not only to reduce carbon emission per unit of GDP, but also to consume increased amounts of non-fossil energy. The carbon emission allowance is a new type of financial asset in each Chinese province and city that also affects individual firms. This paper attempts to examine the allocative efficiency of carbon emission reduction and non-fossil energy consumption by employing a zero sum gains data envelopment analysis (ZSG-DEA) model, given the premise of fixed CO2 emissions as well as non-fossil energy consumption. In making its forecasts, the paper optimizes allocative efficiency in 2020 using 2010 economic and carbon emission data from 30 provinces and cities across China as its baseline. An efficient allocation scheme is achieved for all the provinces and cities using the ZSG-DEA model through five iterative calculations.

Keyword ：

iteration carbon emission allowance zero sum gains data envelopment analysis (ZSG-DEA) non-fossil fuels efficiency

Cite：

Copy from the list or Export to your reference management。

Abstract ：

GPS common-view is the main method for the time and frequency remote transfer currently, of which the uncertainty can reach several nanoseconds. The GPS P3 code receiver has recently become one of the international research hotspots, which can improve the comparison accuracy by revising ionosphere delay real-time values observed at two frequencies. In order to test the accuracy of the GPS Common-view comparison system, two EURO-160 receivers and a SEPT POLARX2 receiver constitutes the zero-baseline Common-view comparison experiment. The test result of experiment shows that the common-view system can reach the accuracy of 2 similar to 3ns, which is better than the single-frequency receivers. This system can present better time & frequency transfer services.

Keyword ：

P3 code receiver time and frequency comparison zero-baseline GPS common-view

Cite：

Copy from the list or Export to your reference management。