Existing surveys on stock market prediction often focus traditional machine learning methods instead of deep methods. This motivates us to provide a structured and comprehensive overview the research prediction. We present four elaborated subtasks propose novel taxonomy summarize state-of-the-art models based neural networks. In addition, we also detailed statistics datasets evaluation metrics commonly used in market. Finally, point out several future directions by sharing some new perspectives

Natural Language Processing (NLP) demonstrates a great potential to support financial decision-making by analyzing the text from social media or news outlets. In this work, we build platform study NLP-aided stock auto-trading algorithms systematically. contrast previous our is characterized three features: (1) We provide for each specific stock. (2) various factors (3) evaluate performance more financial-relevant metrics. Such design allows us develop and in realistic setting. addition...

With the use of electrolytic Cu powder, Zr Si powder and nickel-coated B4C as cladding powders, in-situ synthesized ZrB2-SiC reinforced copper matrix composite coatings were prepared by laser on surface substrate to improve hardness wear resistance. Under condition a energy density at 60 kJ/cm2, macroscopic coating was continuously flat. The microstructure phase analyzed means XRD SEM. reinforcements with nano-scale particle micron-scale needle-like structures in coating, content...

Laser additive manufacturing is an advanced material preparation technology, which has been widely used to prepare various materials, such as polymers, metals, ceramics and composites. Zirconium diboride (ZrB2) reinforced copper composite was fabricated using laser direct energy deposition technology. The microstructure phase composition of the were analyzed, influence density on mechanical properties materials discussed. results showed that needle-like ZrB2 ceramic reinforcement...

One of the significant challenges in reinforcement learning (RL) when dealing with noise is estimating latent states from observations. Causality provides rigorous theoretical support for ensuring that underlying can be uniquely recovered through identifiability. Consequently, some existing work focuses on establishing identifiability a causal perspective to aid design algorithms. However, these results are often derived purely viewpoint, which may overlook specific RL context. We revisit...

Accurately predicting stock returns is crucial for effective portfolio management. However, existing methods often overlook a fundamental issue in the market, namely, distribution shifts, making them less practical future markets or newly listed stocks. This study introduces novel approach to address this challenge by focusing on acquisition of invariant features across various environments, thereby enhancing robustness against shifts. Specifically, we present InvariantStock, designed...

Financial event extraction enables the of comprehensive and accurate information about financial events from documents. This paper explores current methods for extracting at document level, which often involve custom-designed networks processes. We question whether such extensive efforts are truly necessary this task. Our research is motivated by recent developments in generative extraction, have shown success sentence-level but yet to be explored document-level extraction. To fill gap, we...

Gadolinia (Gd2O3) is potentially attractive as a dispersive phase for copper matrix composites due to its excellent thermodynamic stability. In this paper, series of 1.5 vol% nano-Gd2O3/Cu were prepared via an internal oxidation method followed by powder metallurgy in the temperature range 1123–1223 K with holding time 5–60 min. The effects processing parameters on microstructure and properties analyzed. results showed that tensile strength conductivity composite have strong link...

Ti3AlC2 presents a hexagonal layered crystal structure and bridges the gap between metallic ceramic properties, Gadolinia (Gd2O3) has excellent thermodynamic stability, which make them potentially attractive as dispersive phases for Cu matrix composites. In this paper, Cu@Ti3AlC2-Gd2O3/Cu composites, Ti3AlC2-Gd2O3/Cu Gd2O3/Cu composites were prepared by electroless plating, internal oxidation, vacuum hot press sintering. The microstructure effect of plating on properties discussed. results...

Identifying posts of high financial quality from opinions is extraordinary significance for investors. Hence, this paper focuses on evaluating the rationales amateur investors (ERAI) in a shared task, and we present our solutions. The pairwise comparison task aims at extracting post that will trigger higher MPP ML values pairs posts. goal unsupervised ranking to find top 10% with values. We initially model as text classification regression problems. then propose multi-learning approach...

Identifying unexpected domain-shifted instances in natural language processing is crucial real-world applications. Previous works identify the out-of-distribution (OOD) instance by leveraging a single global feature embedding to represent sentence, which cannot characterize subtle OOD patterns well. Another major challenge current methods face learning effective low-dimensional sentence representations hard that are semantically similar in-distribution (ID) data. In this paper, we propose...

Natural Language Processing(NLP) demonstrates a great potential to support financial decision-making by analyzing the text from social media or news outlets. In this work, we build platform study NLP-aided stock auto-trading algorithms systematically. contrast previous our is characterized three features: (1) We provide for each specific stock. (2) various factors (3) evaluate performance more financial-relevant metrics. Such design allows us develop and in realistic setting. addition...

Evaluating the Rationales of Amateur Investors (ERAI) is a task about mining expert-like viewpoints from social media. This paper summarizes our solutions to ERAI shared task, which co-located with FinNLP workshop at EMNLP 2022. There are 2 sub-tasks in ERAI. Sub-task 1 pair-wised comparison where we propose BERT-based pre-trained model projecting opinion pairs common space for classification. an unsupervised learning ranking opinions’ maximal potential profit (MPP) and loss (ML), leverages...