An actor-critic-based portfolio investment method inspired by benefit-risk optimization

Tang, Lili

doi:10.1177/1748301818779059

Cited by 9 publications

(4 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…refined log-optimal strategy and combined it with reinforcement learning [12]. Lili Tang proposed a model-based actor-critic algorithm under uncertain environment is proposed, where the optimal value function is obtained by iteration on the basis of the constrained risk range and a limited number of funds [13]. David W. Lu implemented in Long Short Term Memory (LSTM) recurrent structures with Reinforcement Learning or Evolution Strategies acting as agents The robustness and feasibility of the system is verified on GBPUSD trading [14].…”

Section: Discussionmentioning

confidence: 99%

Adversarial Deep Reinforcement Learning in Portfolio Management

Liang,

Chen,

Zhu

et al. 2018

Preprint

View full text Add to dashboard Cite

In this paper, we implement three state-of-art continuous reinforcement learning algorithms, Deep Deterministic Policy Gradient (DDPG), Proximal Policy Optimization (PPO) and Policy Gradient (PG)in portfolio management. All of them are widely-used in game playing and robot control. What's more, PPO has appealing theoretical propeties which is hopefully potential in portfolio management. We present the performances of them under different settings, including different learning rates, objective functions, feature combinations, in order to provide insights for parameters tuning, features selection and data preparation. We also conduct intensive experiments in China Stock market and show that PG is more desirable in financial market than DDPG and PPO, although both of them are more advanced. What's more, we propose a so called Adversarial Training method and show that it can greatly improve the training efficiency and significantly promote average daily return and sharpe ratio in back test. Based on this new modification, our experiments results show that our agent based on Policy Gradient can outperform UCRP.

show abstract

Section: Discussionmentioning

confidence: 99%

Adversarial Deep Reinforcement Learning in Portfolio Management

Liang,

Chen,

Zhu

et al. 2018

Preprint

View full text Add to dashboard Cite

show abstract

“…However, the lack of cathepsin C cannot completely eliminate the production of IL‐1β (Campden & Zhang, 2019). Cathepsin L may compensate for cathepsin B in some cases and play an indirect role in the activation of the NLRP3 inflammasome (Tang, 2018). Cathepsin S can compensate for the loss of other cathepsins in the activation of NLRP3 inflammasome.…”

Section: Nlrp3 Inflammasome Activation Pathwaysmentioning

confidence: 99%

The NLRP3 inflammasome: Multiple activation pathways and its role in primary cells during ventricular remodeling

Yan

Yang

et al. 2021

Journal Cellular Physiology

View full text Add to dashboard Cite

Inflammasomes are a group of multiprotein signaling complexes located in the cytoplasm. Several inflammasomes have been identified, including NLRP1, NLRP2, NLRP3, AIM2, and NLRC4. Among them, NLRP3 was investigated in most detail, and it was reported that it can be activated by many different stimuli. Increased NLRP3 protein expression and inflammasome assembly lead to caspase‐1 mediated maturation and release of IL‐1β, which triggers inflammation and pyroptosis. The activation of the NLRP3 inflammasome has been widely reported in studies of tumors and neurological diseases, but relatively few studies on the cardiovascular system. Ventricular remodeling (VR) is an important factor contributing to heart failure (HF) after myocardial infarction (MI). Consequently, delaying VR is of great significance for improving heart function. Studies have shown that the NLRP3 inflammasome plays an essential role in the process of VR. Here, we reviewed the latest studies on the activation pathway of the NLRP3 inflammasome, focusing on the effects of the NLRP3 inflammasome in primary cells during VR, and finally discuss future research directions in this field.

show abstract

“…That is, RL can directly output trade positions, bypassing the explicit forecasting step by the trial-and-error interaction with the financial environment. Many existing RL methods get promising results by focusing on various technologies to extract richer representation, e.g., by modelbased learning (Tang 2018;Yu et al 2019), by adversarial learning (Liang et al 2018), or by state augmentation (Ye et al 2020). However, these RL algorithms assume that portfolio weights can change immediately at the last price once an order is placed.…”

Section: Introductionmentioning

confidence: 99%

Commission Fee is not Enough: A Hierarchical Reinforced Framework for Portfolio Management

Wang

Wei

et al. 2021

AAAI

View full text Add to dashboard Cite

Portfolio management via reinforcement learning is at the forefront of fintech research, which explores how to optimally reallocate a fund into different financial assets over the long term by trial-and-error. Existing methods are impractical since they usually assume each reallocation can be finished immediately and thus ignoring the price slippage as part of the trading cost. To address these issues, we propose a hierarchical reinforced stock trading system for portfolio management (HRPM). Concretely, we decompose the trading process into a hierarchy of portfolio management over trade execution and train the corresponding policies. The high-level policy gives portfolio weights at a lower frequency to maximize the long-term profit and invokes the low-level policy to sell or buy the corresponding shares within a short time window at a higher frequency to minimize the trading cost. We train two levels of policies via a pre-training scheme and an iterative training scheme for data efficiency. Extensive experimental results in the U.S. market and the China market demonstrate that HRPM achieves significant improvement against many state-of-the-art approaches.

show abstract

An actor-critic-based portfolio investment method inspired by benefit-risk optimization

Cited by 9 publications

References 22 publications

Adversarial Deep Reinforcement Learning in Portfolio Management

Adversarial Deep Reinforcement Learning in Portfolio Management

The NLRP3 inflammasome: Multiple activation pathways and its role in primary cells during ventricular remodeling

Commission Fee is not Enough: A Hierarchical Reinforced Framework for Portfolio Management

Contact Info

Product

Resources

About