• Journal of Intelligence and Information Systems
• /
• v.16 no.3
• /
• pp.77-97
• /
• 2010

Market timing is an investment strategy which is used for obtaining excessive return from financial market. In general, detection of market timing means determining when to buy and sell to get excess return from trading. In many market timing systems, trading rules have been used as an engine to generate signals for trade. On the other hand, some researchers proposed the rough set analysis as a proper tool for market timing because it does not generate a signal for trade when the pattern of the market is uncertain by using the control function. The data for the rough set analysis should be discretized of numeric value because the rough set only accepts categorical data for analysis. Discretization searches for proper "cuts" for numeric data that determine intervals. All values that lie within each interval are transformed into same value. In general, there are four methods for data discretization in rough set analysis including equal frequency scaling, expert's knowledge-based discretization, minimum entropy scaling, and na$\ddot{i}$ve and Boolean reasoning-based discretization. Equal frequency scaling fixes a number of intervals and examines the histogram of each variable, then determines cuts so that approximately the same number of samples fall into each of the intervals. Expert's knowledge-based discretization determines cuts according to knowledge of domain experts through literature review or interview with experts. Minimum entropy scaling implements the algorithm based on recursively partitioning the value set of each variable so that a local measure of entropy is optimized. Na$\ddot{i}$ve and Booleanreasoning-based discretization searches categorical values by using Na$\ddot{i}$ve scaling the data, then finds the optimized dicretization thresholds through Boolean reasoning. Although the rough set analysis is promising for market timing, there is little research on the impact of the various data discretization methods on performance from trading using the rough set analysis. In this study, we compare stock market timing models using rough set analysis with various data discretization methods. The research data used in this study are the KOSPI 200 from May 1996 to October 1998. KOSPI 200 is the underlying index of the KOSPI 200 futures which is the first derivative instrument in the Korean stock market. The KOSPI 200 is a market value weighted index which consists of 200 stocks selected by criteria on liquidity and their status in corresponding industry including manufacturing, construction, communication, electricity and gas, distribution and services, and financing. The total number of samples is 660 trading days. In addition, this study uses popular technical indicators as independent variables. The experimental results show that the most profitable method for the training sample is the na$\ddot{i}$ve and Boolean reasoning but the expert's knowledge-based discretization is the most profitable method for the validation sample. In addition, the expert's knowledge-based discretization produced robust performance for both of training and validation sample. We also compared rough set analysis and decision tree. This study experimented C4.5 for the comparison purpose. The results show that rough set analysis with expert's knowledge-based discretization produced more profitable rules than C4.5.

• Tuberculosis and Respiratory Diseases
• /
• v.41 no.4
• /
• pp.379-388
• /
• 1994

Background : In sleep apnea syndrome, arterial oxygen saturation($SaO_2$) decreases at a variable rate and to a variable degree for a given apneic period from patient to patient, and various kinds of cardiac arrythmia are known to occur. Factors supposed to affect arterial oxygen desaturation during apnea are duration of apnea, lung voulume at which apnea occurs, and oxygen consumption rate of the subject. The lung serves as preferential oxygen source during apnea, and there have been many reports related with the influence of lung volume on $SaO_2$ during apnea, but there are few, if any, studies about the influence of oxygen consumption rate of an individual on $SaO_2$ during breath holding or about the profile of arterial oxygen resaturation after breathing resumed. Methods : To investigate the changes of $SaO_2$ and heart rate(HR) during breath holding(BH) and rebreathing(RB) and to evaluate the physiologic factors responsible for the changes, lung volume measurements, and arterial blood gas analyses were performed in 17 healthy subjects. Nasal airflow by thermistor, $SaO_2$ by pulse oxymeter and ECG tracing were recorded on Polygraph(TA 4000, Gould, U.S.A.) during voluntary BH & RB at total lung capacity(TLC), at functional residual capacity(FRC) and at residual volume(RV), respectively, for the study subjects. Each subject's basal metabolic rate(BMR) was assumed on Harris-Benedict equation. Results: The time needed for $SaO_2$ to drop 2% from the basal level during breath holding(T2%) were $70.1{\pm}14.2$ sec(mean${\pm}$standard deviation) at TLC, $44.0{\pm}11.6$ sec at FRC, and $33.2{\pm}11.1$ sec at RV(TLC vs. FRC, p<0.05; FRC vs. RV, p<0.05). On rebreathing after $SaO_2$ decreased 2%, further decrement in $SaO_2$ was observed and it was significantly greater at RV($4.3{\pm}2.1%$) than at TLC($1.4{\pm}1.0%$)(p<0.05) or at FRC($1.9{\pm}1.4%$)(p<0.05). The time required for $SaO_2$ to return to the basal level after RB(Tr) at TLC was not significantly different from those at FRC or at RV. T2% had no significant correlation either with lung volumes or with BMR respectively. On the other hand, T2% had significant correlation with TLC/BMR(r=0.693, p<0.01) and FRC/BMR (r=0.615, p<0.025) but not with RV/BMR(r=0.227, p>0.05). The differences between maximal and minimal HR(${\Delta}HR$) during the BH-RB manuever were $27.5{\pm}9.2/min$ at TLC, $26.4{\pm}14.0/min$ at RV, and $19.1{\pm}6.0/min$ at FRC which was significantly smaller than those at TLC(p<0.05) or at RV(p<0.05). The mean difference of 5 p-p intervals before and after RB were $0.8{\pm}0.10$ sec and $0.72{\pm}0.09$ sec at TLC(p<0.001), $0.82{\pm}0.11$ sec and $0.73{\pm}0.09$ sec at FRC(p<0.025), and $0.77{\pm}0.09$ sec and $0.72{\pm}0.09$ sec at RV(p<0.05). Conclusion Healthy subjects showed arterial desaturation of various rates and extent during breath holding at different lung volumes. When breath held at lung volume greater than FRC, the rate of arterial desaturation significantly correlated with lung volume/basal metabolic rate, but when breath held at RV, the rate of arterial desaturation did not correlate linearly with RV/BMR. Sinus arrythmias occurred during breath holding and rebreathing manuever irrespective of the size of the lung volume at which breath holding started, and the amount of change was smallest when breath held at FRC and the change in vagal tone induced by alteration in respiratory movement might be the major responsible factor for the sinus arrythmia.

• Tuberculosis and Respiratory Diseases
• /
• v.45 no.1
• /
• pp.140-152
• /
• 1998

Background: Prone position improves oxygenation in patients with ARDS probably by reducing shunt Reduction of shunt in prone position is thought to be effected by lowering of the critical opening pressure (COP) of the dorsal lung because the pleural pressure becomes less positive in prone position compared to supine position. It can then be assumed that prone position would bring about greater improvement in oxygenation when PEEP applied in supine position is just beneath COP than when PEEP is above COP. Hemodynamically, prone position is expected to attenuate the lifting of cardiac fossa induced by PEEP. Based on these backgrounds, we investigated whether the effect of prone position on oxygenation differs in magnitude according to the level of PEEP applied in supine position, and whether impaired cardiac output in supine position by PEEP can be restored in prone position. Methods: In seven mongrel dogs, $PaO_2/F_1O_2$(P/F) was measured in supine position and at prone position 30 min. Cardiac output (CO), stroke volume (SV), pulse rate (PR), and pulmonary artery occlusion pressure (PAOP) were measured in supine position, at prone position 5 min, and at prone position 30 min. After ARDS was established with warmed saline lavage(P/F ratio $134{\pm}72$ mm Hg), inflection point was measured by constant flow method($6.6{\pm}1.4cm$ $H_2O$), and the above variables were measured in supine and prone positions under the application of Low PEEP($5.0{\pm}1.2cm$ $H_2O$), and Optimal PEEP($9.0{\pm}1.2cm$ $H_2O$)(2 cm $H_2O$ below and above the inflection point, respectively) consecutively. Results : P/F ratio in supine position was $195{\pm}112$ mm Hg at Low PEEP and $466{\pm}63$ mm Hg at Optimal PEEP(p=0.003). Net increase of P/F ratio at prone position 30 min, however, was far greater at Low PEEP($205{\pm}90$ mm Hg) than at Optimal PEEP($33{\pm}33$ mm Hg)(p=0.009). Compared to CO in supine position at Optimal PEEP($2.4{\pm}0.5$ L/min), CO in prone improved to $3.4{\pm}0.6$ L/min at prone position 5 min (p=0.0180) and $3.6{\pm}0.7$ L/min at prone position 30 min (p=0.0180). Improvement in CO was attributable to the increase in SV: $14{\pm}2$ ml in supine position, $20{\pm}2$ ml at prone position 5 min (p=0.0180), and $21{\pm}2$ ml at prone position 30 min (p=0.0180), but not to change in PR or PAOP. When the dogs were turned to supine position again, MAP ($92{\pm}23$ mm Hg, p=0.009), CO ($2.4{\pm}0.5$ L/min, p=0.0277) and SV ($14{\pm}1$ ml, p=0.0277) were all decreased compared to prone position 30 min. Conclusion: Prone position in a dog with saline-lavaged acute lung injury appeared to augment the effect of relatively low PEEP on oxygenation, and also attenuate the adverse hemodynamic effect of relatively high PEEP. These findings suggest that a PEEP lower than Optimal PEEP can be adopted in prone position to achieve the goal of alveolar recruitment in ARDS avoiding the hemodynamic complications of a higher PEEP at the same time.