The Mahalanobis-Taguchi (MT) method helps perform anomaly detection or pattern recognition. This method creates a unit space wherein the normal group forms a homogeneous population, such that whether an individual belongs to the unit space is detected as an anomaly. However, when anticipating the homogeneity of abnormal individuals, which we refer to as “known anomaly,” we can define the unit space for them as well as normal. Any individual that does not belong to any unit space is likely to be an “unknown anomaly” after multiple unit spaces are set accordingly. In this study, we propose two novel analyses to detect and classify any “unknown anomaly” and “known anomaly” within the MT framework. Through Monte Carlo simulation and data analysis, we demonstrate that the proposed procedures can appropriately detect and classify two types of abnormal individuals. We focus not only on “Supervised Learning” using the training data labeled “normal” and “known anomaly,” but also on “Semi-Supervised Learning” using labeled and unlabeled data. We introduce “Semi-Supervised Learning” as a parameter estimation method in the proposed procedures, which confirms its contribution to reducing the labeling cost by engineers.
The Mahalanobis–Taguchi (MT) method is a multivariate analysis method that addresses problems such as pattern recognition and anomaly detection wherein outliers from data groups are detected. The MT method is applied to various other fields (e.g., medical examination, corporate bankruptcy discrimination, and employee turnover discrimination). At the time of application, the effectiveness of the MT method for scale data, other than continuous variables, has not been clarified. The MT method using the polyserial and polychoric correlation (MTP) method, which is an MT method using the Mahalanobis distance calculated by Pearson, polyserial and polychoric correlation coefficients, is proposed. Correlations between continuous variables are calculated using Pearson’s correlation coefficient, continuous and ordinal-scale variables are calculated using the polyserial correlation coefficient, and ordinal-scale variables are calculated using the polychoric correlation coefficient. Through simulation with artificial data, it was confirmed that the anomaly detection accuracy of the MT method decreased with respect to the ordinal scale and the correlation weakening by ordinal scaling can be eliminated using the polyserial and polychoric correlation coefficients. The results of this study indicate that the abnormality discrimination accuracy of the MTP method exceeds that of the MT method, although in a limited environment.
Hospitals play an important role in maintaining social infrastructure and are therefore required to perform emergency response work in addition to their daily routine work in the event of a natural disaster or pandemic. Thus, hospitals regularly conduct disaster exercises and education to raise the crisis awareness of medical staff, preparing for a quick response during a disaster.
However, many hospitals implement disaster exercises only a few times per year. Moreover, even if the exercises are performed, it is often difficult to impart the same feeling of tension as can be felt during a disaster. Therefore, it is difficult to encourage medical staff to develop their crisis awareness and actions, which may inhibit preparation for disasters.
This study proposed a mechanism that results in a change in crisis awareness and action with reference to past research and subsequently developed a questionnaire based on this mechanism. The questionnaire was administered to the staff of a certain base disaster hospital to clarify the reasons for crisis awareness and action changes in response to disasters. The results of this study will make it possible to design both disaster exercises and education that promote crisis awareness and action changes among medical staff.
In this study, crisis awareness is defined as the awareness that triggers action in response to an event after it is recognized through proactive information gathering and disaster response, and a change in action is defined as an individual healthcare staff member’s effort to solve the problems faced by the hospital.
This study examines the effects of implementing an Occupational Health and SafetyManagement System (OHSMS) and obtaining OHSMS certification. Insofar as the previousresearch on OHSMS certification has been largely based on surveys of certifiedcompanies only, the effect of implementing an OHSMS and the effect of obtainingOHSMS certification have generally been confounded. In this study, we investigateboth companies that have implemented an OHSMS with certification and companies thathave implemented an OHSMS without certification in order to comparetheir Occupational Health and Safety (OHS) performance. In addition, whileprevious studies capture the effects of an OHSMS discretely, this study identifiesinterrelated effects based on the proposed causal model. Results of the studyshow that companies that have sought and received OHSMS certification realize greaterbenefits in a variety of areas, including a reduction in the risk ofcompensation for damage, improvements in operational performance, and enhancedcompany sustainability, as compared to non-certified companies. Suchclarification of the differences between certified companies and non-certifiedcompanies should encourage more companies to pursue OHSMS certification.
Load-sharing is one way to improve the reliability of amulti-unit system. However, it also increases maintenance costs. Considerationof the trade-off between improving reliability and reducing maintenance costsis complicated by the fact that workload allocation can greatly affect the unitdeterioration rate. Decision making regarding maintenance and load-sharing musttherefore take into account the complex and dynamic interactions between unitdeterioration and workload allocation. This research considers condition-basedmaintenance for a system consisting of two identical units. An optimal joint maintenanceand load-sharing policy with flexible load-sharing is presented for two-unitdeteriorating systems under a constant total workload. The underlyingdeterioration process of the system, which depends on the workload allocation,is described by a continuous stochastic process. The deterioration state isobserved at the beginning of equally spaced time periods. The operator caneither continue operating the system for one more period with a certain loaddistribution ratio or initiate maintenance on either or both units on the basisof the deterioration state of the system. The decision-making problem isformulated as a Markov decision process that minimizes the total expected cost overan infinite horizon. The properties of the resulting optimal decision policies wereanalyzed, and a set of sufficient conditions for a monotone policy wereidentified.
Age-replacement is one of the most commonly usedmaintenance policies based on preventive action to prevent system failure. Inmany age-replacement settings, it is assumed that the maintenance time can beignored since it is short compared with operation time. However, maintenance,especially corrective maintenance, may take a non-negligible amount of time dueto various factors such as ordering replacement parts, equipping themaintenance team, and transporting the maintenance equipment. Neglectingmaintenance time when optimizing maintenance policies for multi-unit systems maythus lead to sub-optimal or even incorrect solutions to the problem, resultingin higher maintenance costs. We present an optimal age-replacement policy forsystems consisting of multiple independent units connected in series. Thefailure of any one of the units causes the entire system to fail, which isimmediately detected. We formulated a model that integrates both preventive andcorrective maintenance times for determining an optimal preventive replacementinterval, minimizing the expected long run cost per unit time. The sufficientconditions for the existence and uniqueness of the corresponding optimalsolution are derived. The performance of the proposed maintenance policy wasevaluated by comparing it with a conventional one that does not take intoaccount maintenance time. Numerical examples illustrate the effects of downtimecost and maintenance time on the proposed maintenance policy.