检索结果-内蒙古大学图书馆

33rd ACM International Conference on Information and Knowledge Management (CIKM)

作者： Fang, Yupeng Liu, Ruirui Huang, Huichou Zhao, Peilin Wu, Qingyao South China Univ Technol Guangzhou Peoples R China Brunel Univ London Kings Coll London London England City Univ Hong Kong Bayescien Technol Hong Kong Peoples R China Tencent AI Lab Shenzhen Peoples R China Peng Cheng Lab Shenzhen Peoples R China

ISBN: (纸本)9798400704369

Missing values and unreleased figures are common but highly important for backtesting and real-time analysis in the financial industry, yet underexploited in the existing literature. In this paper, we focus on the issue of empirical asset pricing, where the cross-section of future asset returns is a function of lagged firm characteristics that vary in time frequencies and missing ratios. Most of the existing imputation methods cannot fully capture the complex and evolving spatio-temporal relations among firm-level characteristics. In particular, these methods fail to explicitly consider the spatial relations and feature structure in the stock network where we have to process granular data of thousands of stocks and hundreds of characteristics for each stock. To address these challenges, we propose a spatio-temporal diffusion model (STDM) that gradually recovers the masked financial data conditioning on high-dimensional stock-and-characteristics historical data. We propose characteristic-specific projection to construct characteristic-level features at both ends of the STDM, meanwhile maintaining firm-level features in the middle of the STDM to largely reduce the computational memory. Moreover, along with the temporal attention, we design a spatial graph convolutional network, making it computationally efficient and effective to learn time-varying spatio-temporal interdependence across firms. We further employ an implicit sampler that greatly accelerates the inference procedure so that the STDM is able to produce high-quality point and density estimates of missing and real-time firm characteristics within a few steps. We evaluate our model on the most comprehensive open-source dataset 'OSAP' and generate state-of-the-art performance in extensive experiments.

关键词： Missing Value Imputation Real-Time Nowcasting Diffusion Model financial data processing

来源：评论

学校读者我要写书评

暂无评论

AdaFT: An efficient domain-adaptive fine-tuning framework for sentiment analysis in chinese financial texts

引用

APPLIED INTELLIGENCE 2025年第7期55卷 1-23页

作者： Yan, Guofeng Peng, Kuashuai Wang, Yongfeng Tan, Hengliang Du, Jiao Wu, Heng Guangzhou Univ Sch Comp Sci & Cyber Engn Guangzhou 510006 Guangdong Peoples R China Univ La Verne Coll Art & Sci Comp Sci & Comp Engn Dept 1950 3rd St La Verne CA 91724 USA

Given the prevalence of pre-trained language models (PLMs) within the field of natural language processing, it has become evident that the conventional two-stage approach of 'pre-training'-then-'fine-tuning' consistently yields commendable outcomes. Nevertheless, most publicly accessible PLMs are pre-trained on extensive, general-purpose datasets, thereby failing to address the substantial domain dissimilarity between the source and target data. This discrepancy has significant implications for the adaptability of PLMs to specific domains. To address this issue, our study proposes AdaFT, an efficient domain-adaptive fine-tuning framework that seeks to enhance the traditional fine-tuning process, thus bridging the gap between the source and target domains. This is particularly beneficial for enabling PLMs to better align with the specialized context of the Chinese financial domain. In contrast to the standard two-stage paradigm, AdaFT incorporates two additional stages: 'multi-task further pre-training' and 'multi-model parameter fusion.' In the first phase, the PLM undergoes a rapid, multi-task, parallel learning process, which effectively augments its proficiency in Chinese financial domain-related tasks. In the subsequent stage, we introduce an adaptive multi-model parameter fusion (AdaMFusion) strategy to amalgamate the knowledge acquired from the extended pre-training. To efficiently allocate weights for AdaMFusion, we have developed a local search algorithm with a decreasing step length, i.e., Local Search with Decreasing Step size (LSDS). The combination of AdaMFusion and LSDS algorithm strikes a balance between efficiency and performance, making it suitable for most scenarios. We also find that the optimal weighting factor assigned to a model to be fused is positively correlated with the performance improvement of that model on the target task after further pre-training. We demonstrate that further pre-training is generally effective, and further pre-training

关键词： Domain adaptation Pre-trained language model Fine-tuning financial data processing Weight allocation

来源：评论

学校读者我要写书评

暂无评论

Optimizing Accounting Informatization through Simultaneous Multi-Tasking across Edge and Cloud Devices using Hybrid Machine Learning Models

引用

JOURNAL OF GRID COMPUTING 2024年第1期22卷 12-12页

作者： Yang, Xiaofeng Chongqing Ind Polytech Coll Sch Econ & Management Chongqing 401120 Peoples R China

Accounting informatization is a crucial component of enterprise informatization, significantly impacting operational efficiency in accounting and finance. Advances in information technology have introduced automation techniques that accelerate the processing of accounting information cost-effectively. Integrating artificial intelligence, cloud computing, and edge computing is pivotal in streamlining and optimizing these processes. Traditionally, accounting informatization relied on system servers and local storage for data processing. However, the era of big data necessitates a shift to cloud computing frameworks for efficient data storage and processing. Despite the advantages of cloud storage, concerns arise regarding data security and the substantial data transactions between the cloud and source devices. To address these challenges, this research proposes a novel algorithm, Heterogeneous Distributed Deep Learning with data Offloading (DDLO) algorithm. DDLO leverages the synergy between edge devices and cloud computing to enhance data processes. Edge computing enables rapid processing of large volumes of data at or near the data collection sites, optimizing day-to-day operations for enterprises. Furthermore, machine learning algorithms at edge devices enhance data processing efficiency, augmenting the computing environment's overall performance. The proposed DDLO algorithm fosters a hybrid machine learning approach for computing joint tasks and multi-tasking in accounting informatization. It enables dynamic resource allocation, allowing selected data or model updates to be offloaded to the cloud for complex tasks. The algorithm's performance is rigorously evaluated using key metrics, including computing time, offloading time, accuracy, and cost levels. By capitalizing on the strengths of edge computing, cloud computing, and artificial intelligence, the DDLO algorithm effectively addresses accounting informatization challenges. It empowers enterprises to process v

关键词： Accounting informatization Hybrid machine learning Edge computing Cloud computing financial data processing Multitasking

来源：评论

学校读者我要写书评

暂无评论

Autoencoder-Enhanced Clustering: A Dimensionality Reduction Approach to financial Time Series

引用

IEEE ACCESS 2024年 12卷 16999-17009页

作者： Cortes, Daniel Gonzalez Onieva, Enrique Lopez, Iker Pastor Trinchera, Laura Wu, Jian NEOMA Business Sch F-76825 Mont St Aignan France Univ Deusto Fac Engn Bilbao 48007 Spain

While Machine Learning significantly boosts the performance of predictive models, its efficacy varies across different data dimensions. It is essential to cluster time series data of similar characteristics, particularly in the financial sector. However, clustering financial time series data poses considerable challenges due to the market's inherent complexity and multidimensionality. To address these issues, our study introduces a novel clustering framework that leverages autoencoders for a compressed yet informative representation of financial time series. We rigorously evaluate our approach through multiple dimensionality reduction and clustering algorithms, applying it to key financial indices, including IBEX-35, CAC-40, DAX-30, S&P 500, and FTSE 100. Our findings consistently demonstrate that incorporating autoencoders significantly enhances the granularity and quality of clustering, effectively isolating distinct categories of financial time series. Our findings carry significant ramifications for the financial industry. By refining clustering methodologies, we set the stage for increasingly accurate financial predictive models, offering valuable insights for optimizing investment strategies and enhancing risk management.

关键词： Time series analysis Principal component analysis Clustering algorithms Memory management Clustering methods Dimensionality reduction Decoding data compression financial services data compression financial data processing neural network applications time series

来源：评论

学校读者我要写书评

暂无评论

Quantum Computational Finance: Quantum Algorithm for Portfolio Optimization

引用

KUNSTLICHE INTELLIGENZ 2024年第4期38卷 327-338页

作者： Rebentrost, Patrick Lloyd, Seth Natl Univ Singapore Ctr Quantum Technol 3 Sci Dr 2 Singapore 117543 Singapore MIT Dept Mech Engn Cambridge MA 02139 USA MIT Res Lab Elect Cambridge MA 02139 USA

We present a quantum algorithm for portfolio optimization. We discuss the market data input of asset prices, the processing of such data via quantum operations, and the output of financially relevant results. Given quantum access to a historical record of asset returns, the algorithm determines the optimal risk-return tradeoff curve and allows one to sample from the optimal portfolio. The algorithm can in principle attain a run time of poly(log(N))\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\textrm{poly}(\log (N))$$\end{document}, where N is the number of assets. Direct classical algorithms for determining the risk-return curve and other properties of the optimal portfolio take time poly(N)\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\textrm{poly}(N)$$\end{document} and we discuss potential quantum speedups in light of efficient classical sampling approaches.

关键词： financial data processing

来源：评论

学校读者我要写书评

暂无评论

Chinese fine-grained financial sentiment analysis with large language models

引用

Neural Computing and Applications 2024年 1-10页

作者： Lan, Yinyu Wu, Yanru Xu, Wang Feng, Weiqiang Zhang, Youhao School of Medicine and Health Harbin Institute of Technology Heilongjiang Harbin150001 China AI Research FinChina Anhui Hefei230031 China

Entity-level fine-grained sentiment analysis in the financial domain is a crucial subtask of sentiment analysis and currently faces numerous challenges. The primary challenge stems from the need for more high-quality and large-scale annotated corpora designed explicitly for financial text sentiment analysis, which, in turn, limits the availability of data necessary for developing effective text processing techniques. Recent advancements in large language models (LLMs) have yielded remarkable performance in natural language processing tasks, primarily centered around language pattern matching. In this paper, we propose a novel and extensive Chinese fine-grained financial sentiment analysis dataset, FinChina SA, for enterprise early warning. We thoroughly evaluate and experiment with well-known existing open-source LLMs using our dataset. We firmly believe that our dataset will serve as a valuable resource to advance the exploration of real-world financial sentiment analysis tasks, which should be the focus of future research. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： financial data processing

来源：评论

学校读者我要写书评

暂无评论

OPTIMIZING financial data ANALYSIS: A COMPARATIVE STUDY OF PREprocessing TECHNIQUES FOR REGRESSION MODELING OF APPLE INC.'S NET INCOME AND STOCK PRICES

arXiv

引用

arXiv 2025年

作者： Ungar, Kevin Oprean-Stan, Camelia Lucian Blaga University of Sibiu Faculty of Economic Sciences Sibiu Romania

This article presents a comprehensive methodology for processing financial datasets of Apple Inc., encompassing quarterly income and daily stock prices, spanning from March 31, 2009, to December 31, 2023. Leveraging 60 observations for quarterly income and 3774 observations for daily stock prices, sourced from Macrotrends and Yahoo Finance respectively, the study outlines five distinct datasets crafted through varied preprocessing techniques. Through detailed explanations of aggregation, interpolation (linear, polynomial, and cubic spline) and lagged variables methods, the study elucidates the steps taken to transform raw data into analytically rich datasets. Subsequently, the article delves into regression analysis, aiming to decipher which of the five data processing methods best suits capital market analysis, by employing both linear and polynomial regression models on each preprocessed dataset and evaluating their performance using a range of metrics, including cross-validation score, MSE, MAE, RMSE, R-squared, and Adjusted R-squared. The research findings reveal that linear interpolation with polynomial regression emerges as the top-performing method, boasting the lowest validation MSE and MAE values, alongside the highest R-squared and Adjusted R-squared values. © 2025, CC BY.

关键词： financial data processing

来源：评论

学校读者我要写书评

暂无评论

Automating Methods for Validating PV Plant Equipment Labels 52

Automating Methods for Validating PV Plant Equipment Labels

引用

52nd IEEE Photovoltaic Specialist Conference, PVSC 2024

作者： Ranalli, Joseph Hobbs, William B. Penn State Hazleton HazletonPA18202 United States Southern Company BirminghamAL35203 United States

ISBN: (纸本)9781665464260

Operators of utility scale photovoltaic plants can face challenges in accurately matching operational data at various levels with their physical locations in the plant. This may have financial implications due to misdirected maintenance efforts. A method to predict equipment locations using time series data analysis was developed, but previously relied on some manual processing. This paper details efforts to automate the technique for improved efficiency and to facilitate operational deployment. The automated methodology relies on selecting ideal cloud motion vectors by targeting time periods based on their variability score metric. Ten vectors with well-spaced angular directions were selected after filtering based on quality control parameters. After the analysis methodology was run for an entire plant, mislabeled equipment was detected by implementing optimization code for solving the well-known Assignment Problem. Because predicted equipment positions have a circular dependence on their original expected positions, the process was applied iteratively as mislabeled entities were identified to converge to a final assignment of predicted position for each plant component. The results from this algorithmic methodology agree with the validated results from the original manual method. These improvements further the goal of offering the method as a ready-to-use tool for validating the physical locations of equipment within a utility-scale photovoltaic plant. © 2024 IEEE.

关键词： financial data processing

来源：评论

学校读者我要写书评

暂无评论

A Comprehensive Survey on Enhancing Stock Market Predictions through Machine Learning and NLP-Based Sentiment Analysis Integration 4

A Comprehensive Survey on Enhancing Stock Market Predictions...

引用

4th International Conference on Sentiment Analysis and Deep Learning, ICSADL 2025

作者： Sathish, Nandigama Jamalpur, Bhavana School of Computer Science and Artificial Intelligence Sr University Telangana Warangal506371 India

ISBN: (纸本)9798331523923

This research study reviews the integration of Machine Learning (ML) techniques with Natural Language processing (NLP)-based sentiment analysis to improve stock market prediction accuracy by capturing public sentiment from unstructured data sources such as news articles, social media, and financial forums. The study highlights the performance improvements achieved when sentiment data is incorporated into predictive models using ML algorithms, including K-Nearest Neighbors (K-NN), Logistic Regression, Support Vector Machine (SVM), and a hybrid KNN-LR model. A hybrid KNN-LR model combined with sentiment data outperforms individual algorithms, suggesting that combining ML with NLP-based sentiment analysis provides a more nuanced understanding of market behavior. The survey addresses key challenges, such as data quality, processing complexity, and the need for model transparency, while recommending future research on large datasets, advanced NLP techniques, and adaptive machine learning models to enhance accuracy and adaptability in dynamic financial markets. In a rapidly evolving market landscape, sentiment-driven tools will empower investors and analysts in place of theoretical models. © 2025 IEEE.

关键词： financial data processing

来源：评论

学校读者我要写书评

暂无评论

Evaluating Interactive Concept Maps Produced from E-Portfolios 26

Evaluating Interactive Concept Maps Produced from E-Portfoli...

引用

26th International Symposium on Multimedia, ISM 2024

作者： Gantikow, Alexander Isking, Andreas Muller, Wolfgang Libbrecht, Paul Rebholz, Sandra University of Education Weingarten Media Education and Visualization Weingarten Germany Iu International University of Applied Science It and Technology Erfurt Germany Ostbayerische Technische Hochschule Amberg-Weiden Electrical Engineering Media and Computer Science Amberg-Weiden Germany

ISBN: (纸本)9798331511111

Natural Language processing opens new ways to analyze digital works. We are interested in the assessment of e-portfolios which are digital learning logs. Through analysis techniques such as the recognition of topics of a paragraph it is is possible to visualize the topic coverage and content depth. Such visualizations can be useful in the e-portfolio creation and review processes. We have applied these technologies to develop interactive concept maps for exam-relevant e-portfolios and have used them in a real-life setting at university. The experiment has shown that the interactive concept maps have motivated the students and that they have the potential to be useful in the writing process if they are well integrated into the workflow. © 2024 IEEE.

关键词： financial data processing

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：