Some Observations for Portfolio Management Applications of Modern Machine Learning Methods

Park, Jooyoung,Heo, Seongman,Kim, Taehwan,Park, Jeongho,Kim, Jaein,Park, Kyungwook Korean Institute of Intelligent Systems 2016 INTERNATIONAL JOURNAL of FUZZY LOGIC and INTELLIGE Vol.16 No.1

Recently, artificial intelligence has reached the level of top information technologies that will have significant influence over many aspects of our future lifestyles. In particular, in the fields of machine learning technologies for classification and decision-making, there have been a lot of research efforts for solving estimation and control problems that appear in the various kinds of portfolio management problems via data-driven approaches. Note that these modern data-driven approaches, which try to find solutions to the problems based on relevant empirical data rather than mathematical analyses, are useful particularly in practical application domains. In this paper, we consider some applications of modern data-driven machine learning methods for portfolio management problems. More precisely, we apply a simplified version of the sparse Gaussian process (GP) classification method for classifying users' sensitivity with respect to financial risk, and then present two portfolio management issues in which the GP application results can be useful. Experimental results show that the GP applications work well in handling simulated data sets.

Some Observations for Portfolio Management Applications of Modern Machine Learning Methods

Jooyoung Park,Seongman Heo,Taehwan Kim,Jeongho Park,Jaein Kim,Kyungwook Park 한국지능시스템학회 2016 INTERNATIONAL JOURNAL of FUZZY LOGIC and INTELLIGE Vol.16 No.1

Recently, artificial intelligence has reached the level of top information technologies that will have significant influence over many aspects of our future lifestyles. In particular, in the fields of machine learning technologies for classification and decision-making, there have been a lot of research efforts for solving estimation and control problems that appear in the various kinds of portfolio management problems via data-driven approaches. Note that these modern data-driven approaches, which try to find solutions to the problems based on relevant empirical data rather than mathematical analyses, are useful particularly in practical application domains. In this paper, we consider some applications of modern data-driven machine learning methods for portfolio management problems. More precisely, we apply a simplified version of the sparse Gaussian process (GP) classification method for classifying users’ sensitivity with respect to financial risk, and then present two portfolio management issues in which the GP application results can be useful. Experimental results show that the GP applications work well in handling simulated data sets.

The Relationship between Delirium and Statin Use According to Disease Severity in Patients in the Intensive Care Unit

Jun Yong An(Jun Yong An),Jin Young Park(Jin Young Park),Jaehwa Cho(Jaehwa Cho),Hesun Erin Kim(Hesun Erin Kim),Jaesub Park(Jaesub Park),Jooyoung Oh(Jooyoung Oh) 대한정신약물학회 2023 CLINICAL PSYCHOPHARMACOLOGY AND NEUROSCIENCE Vol.21 No.1

Objective: The aim of this study was to investigate the association between the use of statins and the occurrence of delirium in a large cohort of patients in the intensive care unit (ICU), considering disease severity and statin properties. Methods: We obtained clinical and demographical information from 3,604 patients admitted to the ICU from January 2013 to April 2020. This included information on daily statin use and delirium state, as assessed by the Confusion Assessment Method for ICU. We used inverse probability of treatment weighting and categorized the patients into four groups based on the Acute Physiology and Chronic Health Evaluation II score (group 1: 0−10 - mild; group 2: 11−20 - mild to moderate; group 3: 21−30 - moderate to severe; group 4: > 30 - severe). We analyzed the association between the use of statin and the occurrence of delirium in each group, while taking into account the properties of statins. Results: Comparisons between statin and non-statin patient groups revealed that only in group 2, patients who were administered statin showed significantly higher occurrence of delirium (p = 0.004, odds ratio [OR] = 1.58) compared to the patients who did not receive statin. Regardless of whether statins were lipophilic (p = 0.036, OR = 1.47) or hydrophilic (p = 0.032, OR = 1.84), the occurrence of delirium was higher only in patients from group 2. Conclusion: The use of statins may be associated with the increases in the risk of delirium occurrence in patients with mild to moderate disease severity, irrespective of statin properties.

A Modified Approach to Density-Induced Support Vector Data Description

Jooyoung Park,Daesung Kang 한국지능시스템학회 2007 INTERNATIONAL JOURNAL of FUZZY LOGIC and INTELLIGE Vol.7 No.1

The SVDD (support vector data description) is one of the most well-known one-class support vector learning methods, in which one tries the strategy of utilizing balls defined on the feature space in order to distinguish a set of normal data from all other possible abnormal objects. Recently, with the objective of generalizing the SVDD which treats all training data with equal importance, the so-called D-SVDD (density-induced support vector data description) was proposed incorporating the idea that the data in a higher density region are more significant than those in a lower density region. In this paper, we consider the problem of further improving the D-SVDD toward the use of a partial reference set for testing, and propose an LMI (linear matrix inequality)-based optimization approach to solve the improved version of the D-SVDD problems. Our approach utilizes a new class of density-induced distance measures based on the RSDE (reduced set density estimator) along with the LMI-based mathematical formulation in the form of the SDP (semi-definite programming) problems, which can be efficiently solved by interior point methods. The validity of the proposed approach is illustrated via numerical experiments using real data sets.

Controller Design for Fuzzy Systems via Piecewise Quadratic Value Functions

Jooyoung Park,Jongho Kim 한국지능시스템학회 2004 INTERNATIONAL JOURNAL of FUZZY LOGIC and INTELLIGE Vol.4 No.3

This paper concerns controller design for the Takagi-Sugeno (TS) fuzzy systems. The design method proposed in this paper is derived in the framework of the optimal control theory utilizing the piecewise quadratic optimal value functions. The major part of the proposed design procedure consists of solving linear matrix inequalities (LMIs). Since LMIs can be solved efficiently within a given tolerance by the recently developed interior point methods, the design procedure of this paper is useful in practice. A design example is given to illustrate the applicability of the proposed method.

Development and Implication of a Korean Waste Input-Output Table

( Jooyoung Park ),( Hyungwoo Lim ),( Hye Sook Lim ),( Munsol Ju ) 한국폐기물자원순환학회 2022 ISSE 초록집 Vol.2022 No.-

Korea’s waste policy has focused mainly on managing direct waste generation from households, industries, and construction, putting less attention to indirect waste generation driven by final demands. This lack of attention to indirect waste flows may underestimate the impacts of inter-industry relationships and consumption behavioral changes, and therefore lead to missing opportunities for waste reduction. To understand the patterns of direct and indirect waste generation as well as waste management from the consumption perspective, we constructed the first waste input-output table for Korea and analyzed waste footprint of consumption. Korea’s various waste statistics allowed the construction of waste input-output table (WIO) that covers 33 industrial sectors and 69 waste types for the year 2019. Our preliminary results showed that the waste inducement effect was highest in the water supply and waste management sector and the lowest in the financial and insurance sector. Manufacturing sector contributed the most to the generation of industrial waste, particularly basic metals and non-metallic manufacturing, while service sectors contributed mainly to the generation of non-industrial waste. The share of non-combustible waste generation turned out to be considerably high across all sectors, imposing burdens on waste treatment. We will further analyze waste footprint of household consumption and discuss implications of WIO analysis on the waste reduction policy.

Inter-relationships between performance shaping factors for human reliability analysis of nuclear power plants

Park, Jooyoung,Jung, Wondea,Kim, Jonghyun Korean Nuclear Society 2020 Nuclear Engineering and Technology Vol.52 No.1

Performance shaping factors (PSFs) in a human reliability analysis (HRA) are one that may influence human performance in a task. Most currently applicable HRA methods for nuclear power plants (NPPs) use PSFs to highlight human error contributors and to adjust basic human error probabilities (HEPs) that assume nominal conditions of NPPs. Thus far, the effects of PSFs have been treated independently. However, many studies in the fields of psychology and human factors revealed that there may be relationships between PSFs. Therefore, the inter-relationships between PSFs need to be studied to better reflect their effects on operator errors. This study investigates these inter-relationships using two data sources and also suggests a context-based approach to treat the inter-relationships between PSFs. Correlation and factor analyses are performed to investigate the relationship between PSFs. The data sources are event reports of unexpected reactor trips in Korea and an experiment conducted in a simulator featuring a digital control room. Thereafter, context-based approaches based on the result of factor analysis are suggested and the feasibility of the grouped PSFs being treated as a new factor to estimate HEPs is examined using the experimental data.

Ferroelectric (Ba,Sr)TiO3 thin films on Pt/Ti/SiO2/Si substrates by the sol-gel process and evaluation of the intrinsic dead layers

Jooyoung Kim,Gwangseo Park,Jaemoon Park,Kwangwoo Nam 한국물리학회 2005 THE JOURNAL OF THE KOREAN PHYSICAL SOCIETY Vol.47 No.2

Ferroelectric (Ba,Sr)TiO3 (BST) films with a Ba : Sr ratio of 50 : 50 were deposited on Pt/Ti/SiO2/Si substrates by the sol-gel process. The films were spin-coated at 2000 rpm for 30 sec and then pyrolysed for 5 min at a temperature of 350 C. This coating procedure was repeated 3, 4, 5 and 6 times to obtain BST films with different thicknesses. After coating the films with the desired repetition times, the films were finally annealed in a conventional furnace at temperatures ranging from 600 C to 800 C with a 50 C interval in between. The films obtained with an annealing procedure of 750 C were polycrystalline with the presence of an impurity BaCO3 phase. The capacitance and leakage current were measured and used to extract information on the metal-BST interface. With the series capacitance model and modified Schottky emission equation, we will report on the evaluation of dead layer thicknesses in BST films sandwiched between noble metal electrodes.kwhere

One-Class Support Vector Learning and Linear Matrix Inequalities

Park, Jooyoung,Kim, Jinsung,Lee, Hansung,Park, Daihee Korean Institute of Intelligent Systems 2003 INTERNATIONAL JOURNAL of FUZZY LOGIC and INTELLIGE Vol.3 No.1

The SVDD(support vector data description) is one of the most well-known one-class support vector learning methods, in which one tries the strategy of utilizing balls defined on the kernel feature space in order to distinguish a set of normal data from all other possible abnormal objects. The major concern of this paper is to consider the problem of modifying the SVDD into the direction of utilizing ellipsoids instead of balls in order to enable better classification performance. After a brief review about the original SVDD method, this paper establishes a new method utilizing ellipsoids in feature space, and presents a solution in the form of SDP(semi-definite programming) which is an optimization problem based on linear matrix inequalities.

Modern Probabilistic Machine Learning and Control Methods for Portfolio Optimization

Park, Jooyoung,Lim, Jungdong,Lee, Wonbu,Ji, Seunghyun,Sung, Keehoon,Park, Kyungwook Korean Institute of Intelligent Systems 2014 INTERNATIONAL JOURNAL of FUZZY LOGIC and INTELLIGE Vol.14 No.2

Many recent theoretical developments in the field of machine learning and control have rapidly expanded its relevance to a wide variety of applications. In particular, a variety of portfolio optimization problems have recently been considered as a promising application domain for machine learning and control methods. In highly uncertain and stochastic environments, portfolio optimization can be formulated as optimal decision-making problems, and for these types of problems, approaches based on probabilistic machine learning and control methods are particularly pertinent. In this paper, we consider probabilistic machine learning and control based solutions to a couple of portfolio optimization problems. Simulation results show that these solutions work well when applied to real financial market data.