Domain Adaptive Multi-Modality Neural Attention Network for Financial Forecasting

Zhou, Dawei; Zheng, Lecheng; Zhu, Yada; Li, Jianbo; He, Jingrui

doi:10.1145/3366423.3380288

Cited by 28 publications

(15 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, Relative Strength Index was adopted to predict significant stock price changes and got superior experimental results by Kamalov (2019) [39]. In addition, multi-source data was also considered for predicting stock prices [31,5,40]. For example, Zhou, Gao et al (2020) [40] employed multiple heterogeneous data sources, including historical transaction data, technical indicators, stock posts, news and Baidu index, to predict the directions of stock price movements.…”

Section: ) Data Dimensionmentioning

confidence: 99%

“…In fact, apart from machine learning [3], knowledge engineering [4] is also an effective method of artificial intelligence. Moreover, knowledge engineering, especially ontology model, not only has strong knowledge representation ability and reasoning ability, but also can provide adequate explanation, which is very crucial for financial statements fraud detection and risk warning [5]. Therefore, intelligent ontology reasoning may be a powerful tool for financial statements fraud detection.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Finding Misstatement Accounts in Financial Statements through Ontology Reasoning

2024

View full text Add to dashboard Cite

Finding misstatement accounts in financial statements, is a key problem of fraud detection. Potential applications include external audit, internal controls, investment decision and securities market regulation. However, most existing intelligent methods just detect financial statements fraud at the company level, while little research can detect financial statements fraud at the account level. For this, to achieve intelligent fraud detection at the accounts level, an ontology-based fraud detection framework was proposed. To be specific, the proposed framework mainly combines the articulation between different accounts and periods, and 30 financial indicators (ratios) as the knowledge basis of ontology. Notably, with OWL (Ontology Web Language), SWRL (Semantic Web Rule Language) and Proté gé ontology editor, the case study not only completed the fraud detection in a fast and timely manner, but also provided logical explanation and risk warning at the accounts level. This fully shows the great advantages and applicability of the proposed framework in the detection of misstatements accounts. Moreover, the proposed framework is of great significance for timely detection, prevention and response of financial statements fraud. More importantly, the proposed framework opens-up a new direction of using ontology reasoning techniques to find misstatement accounts in financial statements, which provides an interpretable and fine-grained way for fraud detection.

show abstract

Section: ) Data Dimensionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Finding Misstatement Accounts in Financial Statements through Ontology Reasoning

2024

View full text Add to dashboard Cite

show abstract

“…Besides different ML models, there is also a trend to utilize alternative data for improving prediction performance. For instance, economic news [51], frequency of prices [141], social media [134], financial events [30], investment behaviors [21] and weather information [146] have been used as extra information to learn intrinsic pattern of financial assets.…”

Section: Financial Forecastingmentioning

confidence: 99%

Reinforcement Learning for Quantitative Trading

Sun,

Wang,

2021

Preprint

View full text Add to dashboard Cite

Quantitative trading (QT), which refers to the usage of mathematical models and data-driven techniques in analyzing the financial market, has been a popular topic in both academia and financial industry since 1970s. In the last decade, reinforcement learning (RL) has garnered significant interest in many domains such as robotics and video games, owing to its outstanding ability on solving complex sequential decision making problems. RL's impact is pervasive, recently demonstrating its ability to conquer many challenging QT tasks. It is a flourishing research direction to explore RL techniques' potential on QT tasks. This paper aims at providing a comprehensive survey of research efforts on RL-based methods for QT tasks. More concretely, we devise a taxonomy of RL-based QT models, along with a comprehensive summary of the state of the art. Finally, we discuss current challenges and propose future research directions in this exciting field.

show abstract

“…On the contrary, MuFasa constructs an assembled neural networks to learn 𝐾 bandits jointly. Deep neural network in multi-view learning has been well-studied [21,25,43,44,46], to extract useful information among multiple sources, which inspires one of the core ideas of MuFasa. Other variant bandit setting.…”

Section: Related Workmentioning

confidence: 99%

Multi-facet Contextual Bandits: A Neural Network Perspective

Ban,

He,

Cook

2021

Preprint

Self Cite

View full text Add to dashboard Cite

Contextual multi-armed bandit has shown to be an effective tool in recommender systems. In this paper, we study a novel problem of multi-facet bandits involving a group of bandits, each characterizing the users' needs from one unique aspect. In each round, for the given user, we need to select one arm from each bandit, such that the combination of all arms maximizes the final reward. This problem can find immediate applications in E-commerce, healthcare, etc. To address this problem, we propose a novel algorithm, named MuFasa, which utilizes an assembled neural network to jointly learn the underlying reward functions of multiple bandits. It estimates an Upper Confidence Bound (UCB) linked with the expected reward to balance between exploitation and exploration. Under mild assumptions, we provide the regret analysis of Mu-Fasa. It can achieve the near-optimal O ((𝐾 + 1) √𝑇 ) regret bound where 𝐾 is the number of bandits and 𝑇 is the number of played rounds. Furthermore, we conduct extensive experiments to show that MuFasa outperforms strong baselines on real-world data sets.

show abstract

Domain Adaptive Multi-Modality Neural Attention Network for Financial Forecasting

Cited by 28 publications

References 17 publications

Finding Misstatement Accounts in Financial Statements through Ontology Reasoning

Finding Misstatement Accounts in Financial Statements through Ontology Reasoning

Reinforcement Learning for Quantitative Trading

Multi-facet Contextual Bandits: A Neural Network Perspective

Contact Info

Product

Resources

About