arXiv subjects

Ju-Hong Lee

Publications and source records attributed to Ju-Hong Lee.

2 recordsLinked to original sources

arXiv2023

Market-Adaptive Ratio for Portfolio Management

Traditional risk-adjusted returns, such as the Treynor, Sharpe, Sortino, and Information ratios, have been pivotal in portfolio asset allocation, focusing on minimizing risk while maximizing profit. Nevertheless, these metrics often fail to account for the distinct characteristics of bull and bear markets, leading to sub-optimal investment decisions. This paper introduces a novel approach called the Market-adaptive Ratio, which was designed to adjust risk preferences dynamically in response to market conditions. By integrating the $\rho$ parameter, which differentiates between bull and bear markets, this new ratio enables a more adaptive portfolio management strategy. The $\rho$ parameter is derived from historical data and implemented within a reinforcement learning framework, allowing the method to learn and optimize portfolio allocations based on prevailing market trends. Empirical analysis showed that the Market-adaptive Ratio outperformed the Sharpe Ratio by providing more robust risk-adjusted returns tailored to the specific market environment. This advance enhances portfolio performance by aligning investment strategies with the inherent dynamics of bull and bear markets, optimizing risk and return outcomes.

q-fin.PM↗

arXiv2017

Learning Solving Procedure for Artificial Neural Network

It is expected that progress toward true artificial intelligence will be achieved through the emergence of a system that integrates representation learning and complex reasoning (LeCun et al. 2015). In response to this prediction, research has been conducted on implementing the symbolic reasoning of a von Neumann computer in an artificial neural network (Graves et al. 2016; Graves et al. 2014; Reed et al. 2015). However, these studies have many limitations in realizing neural-symbolic integration (Jaeger. 2016). Here, we present a new learning paradigm: a learning solving procedure (LSP) that learns the procedure for solving complex problems. This is not accomplished merely by learning input-output data, but by learning algorithms through a solving procedure that obtains the output as a sequence of tasks for a given input problem. The LSP neural network system not only learns simple problems of addition and multiplication, but also the algorithms of complicated problems, such as complex arithmetic expression, sorting, and Hanoi Tower. To realize this, the LSP neural network structure consists of a deep neural network and long short-term memory, which are recursively combined. Through experimentation, we demonstrate the efficiency and scalability of LSP and its validity as a mechanism of complex reasoning.

cs.AI↗