Search

Your search keyword '"Liu, Chunjiang"' showing total 740 results
Start Over Author "Liu, Chunjiang"
740 results on '"Liu, Chunjiang"'

2. A Three-Phases SFT Hybrid Model Integrated Strong Prior Module and Data Overlap Estimation in the Eduation Context

Author

Chen, Zhangquan, Liu, Chunjiang, and Duan, Haobin

Subjects
Computer Science - Machine Learning, Computer Science - Artificial Intelligence, Computer Science - Computation and Language, I.2.7
Abstract

In this paper, we propose an end-to-end prior-based three-phases supervised fine-tuned model, which is proved more competitive than traditional fine-tuning method. More specifically, our model realizes the structural disassembly and incremental guided output of educational knowledge. To this end, we robustify data classification of three types via a sampler and overlap estimation neural network, and inject the preprocessing datasets into pre-trained model in three batches for LORA fine-tuning. Then, we design a prior module couples system prompt, vector databases, and abstract syntax tree task segmentation. Finally, the compression method and regularization constraint are applied to the prior-based fine-tuned model, followed by text filter at the output end to obtain incremental guided results. Our model represents the first research effort to truly embody the tutor role with the features of abundant educational knowledge, step-by-step incremental guided outputs and non-disclosure of answers. Extensive experiments report that our model also achieves state-of-the-art in code abilities compared to open-source models, reaching an impressive 75.10% on the HumanEval (@pass 1) benchmark. Additionally, our model maintains strong conversational capabilities, with the 13B quantized version achieving scores of 56.34, 50.60, and 45.27 respectively on the MMLU, C-Eval, and AGIEval (5 shot) dialogue evaluation benchmarks., Comment: 9 pages, 2 figures

Published
2024

3. A Community Detection and Graph Neural Network Based Link Prediction Approach for Scientific Literature

Author

Liu, Chunjiang, Han, Yikun, Xu, Haiyun, Yang, Shihan, Wang, Kaidi, and Su, Yongye

Subjects
Computer Science - Social and Information Networks, Computer Science - Artificial Intelligence
Abstract

This study presents a novel approach that synergizes community detection algorithms with various Graph Neural Network (GNN) models to bolster link prediction in scientific literature networks. By integrating the Louvain community detection algorithm into our GNN frameworks, we consistently enhance performance across all models tested. For example, integrating Louvain with the GAT model resulted in an AUC score increase from 0.777 to 0.823, exemplifying the typical improvements observed. Similar gains are noted when Louvain is paired with other GNN architectures, confirming the robustness and effectiveness of incorporating community-level insights. This consistent uplift in performance reflected in our extensive experimentation on bipartite graphs of scientific collaborations and citations highlights the synergistic potential of combining community detection with GNNs to overcome common link prediction challenges such as scalability and resolution limits. Our findings advocate for the integration of community structures as a significant step forward in the predictive accuracy of network science models, offering a comprehensive understanding of scientific collaboration patterns through the lens of advanced machine learning techniques.

Published
2024

4. A Comprehensive Survey on Vector Database: Storage and Retrieval Technique, Challenge

Author

Han, Yikun, Liu, Chunjiang, and Wang, Pengfei

Subjects
Computer Science - Databases, Computer Science - Artificial Intelligence
Abstract

A vector database is used to store high-dimensional data that cannot be characterized by traditional DBMS. Although there are not many articles describing existing or introducing new vector database architectures, the approximate nearest neighbor search problem behind vector databases has been studied for a long time, and considerable related algorithmic articles can be found in the literature. This article attempts to comprehensively review relevant algorithms to provide a general understanding of this booming research area. The basis of our framework categorises these studies by the approach of solving ANNS problem, respectively hash-based, tree-based, graph-based and quantization-based approaches. Then we present an overview of existing challenges for vector databases. Lastly, we sketch how vector databases can be combined with large language models and provide new possibilities.

Published
2023

9. A Three-Phases-LORA Finetuned Hybrid LLM Integrated with Strong Prior Module in the Education Context

Author

Chen, Zhangquan, Liu, Chunjiang, Duan, Haobin, Goos, Gerhard, Series Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Wand, Michael, editor, Malinovská, Kristína, editor, Schmidhuber, Jürgen, editor, and Tetko, Igor V., editor

Published
2024
Full Text
View/download PDF

10. Exploring an FPGA-Based Edge Computing Solution for Smart Manufacturing Monitoring: A Case Study on Droplet Recognition

Author

Liu, Chunjiang, Luo, Ming, IFToMM, Series Editor, Ceccarelli, Marco, Advisory Editor, Corves, Burkhard, Advisory Editor, Glazunov, Victor, Advisory Editor, Hernández, Alfonso, Advisory Editor, Huang, Tian, Advisory Editor, Jauregui Correa, Juan Carlos, Advisory Editor, Takeda, Yukio, Advisory Editor, Agrawal, Sunil K., Advisory Editor, Ball, Andrew D., editor, Ouyang, Huajiang, editor, Sinha, Jyoti K., editor, and Wang, Zuolu, editor

Published
2024
Full Text
View/download PDF

17. Research and experimental verification of lightweight loader rims

Author

Ma Long, Dong Zhikui, Liu Chunjiang, Gao Peng, Liang Ao, Hou Binfeng, Qi Xiangdong, and Jiang Yunhong

Subjects
Construction machinery, Lightweight rims, Fatigue analysis, Response surface construction, Multiobjective optimization, Medicine, Science
Abstract

Abstract The safety performance and structural stiffness of a rim, which is the main load-bearing structure of the loader during operation, influence the overall performance, stability, and braking capabilities of the machine. In the industry, researchers are currently pursuing lightweight and high-strength rims as a primary objective. A low weight not only enhances machinery fuel efficiency but also aligns with societal demands for sustainable development, energy conservation, and emission reduction. In this article, multiobjective optimization analysis on rims composed of three different materials is performed, and the relationships between various optimization parameters and target parameters are established using the results of response surface construction. Multiobjective genetic algorithms are utilized to derive various optimization plans, which are subsequently evaluated through static analysis, fatigue analysis, and weight loss analysis. The final optimization plan is determined based on the calculation results while considering production costs. Field tests are conducted on the optimized rims under various working conditions to verify the test results, evaluate the reliability of the finite element analysis results, and confirm the safety of the optimized rim.

Published
2024
Full Text
View/download PDF

23. Variation in mass-loss rate of foliar litter in relation to climate and litter quality in Eurasian forests: differences among functional groups of litter

Author

Kang, Hongzhang, Berg, Björn, Liu, Chunjiang, and Westman, Carl

Subjects
Forestry, SD1-669.5
Abstract

With a data set of litter decomposition collected by means of literature survey, our objectives are 1) to determine the differences in the variation in the first-year mass loss (%) of leaf litter with regard to climate and litter quality among different functional groups of tree species in Eurasian forests, and 2) to determine the difference in effect of mean annual temperature (°C), annual precipitation (dm), as well as concentration of nitrogen (%), and lignin (%) on first-year mass loss over a wide range in climate and litter quality. The main results are as follows. 1) The significant differences between litter types in the relationships between first-year mass loss and climatic factors plus litter quality revealed clearly different decomposition patterns over the continent. Thus, differences were found between coniferous and broadleaf litter, between deciduous broadleaf and evergreen broadleaf as well as between genera and even within a genus, viz. between deciduous and evergreen Quercus. 2) With a change in a relative unit of climate and litter quality variables, there were clear differences in effects of mean annual temperature, annual precipitation, and nitrogen on first-year mass loss for different functional groups of trees. 3) We identified some broadleaf litter species that decomposed to 100% in one year and thus did not contribute to carbon sequestration in a humus layer. Thus, the variation in pattern of foliar litter decomposition with climate and litter quality across functional groups in Eurasian forests showed different decomposition strategies for litter of different groups and genera.

Published
2009
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results