Liwan Liyanage

A practical inventory control policy using operational statistics

Consider the newsvendor inventory control problem with an ambiguous demand. The traditional approach of separating the parameter estimation and the maximization of the expected profit leads to a suboptimal inventory policy. Operational statistics, introduced in this paper, provides a better solution by integrating the estimation and the optimization tasks.

Women's health during mid-life survey : the use of complementary and alternative medicine by symptomatic women transitioning through menopause in Sydney

Objective:To survey the extent of complementary and alternative medicine (CAM) use among women for the alleviation of menopausal symptoms. Design:A total of 1,296 eligible women aged 45 to 65 years were recruited from three Sydney menopause clinics, general practice clinics, and government agencies between July 2003 and July 2004. Volunteers were invited to complete a 19-item questionnaire covering basic demographics, general health status, use of CAM therapies and products, use of pharmaceuticals, and sources of CAM advice. Results:Of respondents, 53.8% had visited a CAM practitioner and/or used a CAM product during the past year, with 34% using a product only and 5% consulting a practitioner only. The most commonly visited practitioners were naturopaths (7.2%) and acupuncturists (4.8%), whereas the most popular products were soy (25.4%) and evening primrose oil (18.4%). Massage, chiropractic, and nutrition were rated the most effective therapies, and phytoestrogen tablets, evening primrose oil, and black cohosh were deemed the most effective products. Of the 59.9% of respondents currently using prescription or over-the-counter pharmaceuticals, 62.5% reported using CAM products during the past 12 months. Of CAM users 71% had informed their doctor about CAM use, whereas 26.4% of respondents reported their doctor had inquired about CAM use. Conclusions:CAM use by women to alleviate menopausal symptoms is common, with several therapies perceived to be effective. Although a significant proportion of women may use CAM in conjunction with pharmaceuticals, relevant communication between medical practitioners and patients remains inadequate and may expose the patient to potential drug-herb interactions.

Quality improvement for the Campbelltown hospital emergency service

In this study the theory of queueing models has been applied to design an appropriate service facility for the Campbelltown hospital emergency service in order to minimise patients waiting time and the associated running cost. Thus the ultimate goal is to optimize the available resources. A computer program is developed to model the distributions of arrival time, waiting time and service time of the system and to estimate their parameters. These parameters are then used to develop a simulation algorithm that estimates the average waiting time for other parametric changes such as arrival rate and the number of servers in the system. Observing how these parametric changes will effect the expected waiting time of the patients and the expected idle time of the doctors this program can be used to design a more efficient system. Hence this will provide a valuable tool for the managers and doctors in scheduling, in order to optimize the efficiency of the emergency service.

A Data mining algorithm to analyse stock market data using lagged correlation

This paper develops an algorithm for predicting the market direction more accurately when two stocks are strongly correlated to each other with a lag of K number of trading days. The forecasting horizon is the lag; therefore this method is suitable for short term capital gains when the correlation is strong.. This will identify the stocks that are closely related, display the daily price movements and its direction side by side and forecast the direction of the price movement for the dependent stock as well as clearly showing the applicable lag. To test the effectiveness of the method, the most correlated stocks were found and prediction of the direction of the price movements made for 3 different dates for training the model. For each date actual data were then used to verify the accuracy of the prediction. In the testing and verification stage the model predicted the direction of the movement of the stock prices accurately 67% of the time. A generic algorithm is specified so that an automated data mining process can be developed. This algorithm takes into consideration the market-wise analysis performed, varying the lag from a lower limit to an upper limit as specified by the user, calculating the correlation coefficient for each independent stock and all other dependent stocks in the market, selects the pairs of stocks where the correlation coefficients are above a user specified range and lists the stocks data graphically side by side for easy comparison. The primary motivation of this paper is threefold. First, this research examines and analyses the use of market-wide lagged correlation analysis as a forecasting tool. Specifically the ability of one stock to predict the future usually short term future trends of a closely correlated another stock. Second, this paper endeavours to determine the feasibility and practicality of using lagged correlation analysis as a forecasting tool for the individual investor. Finally this paper specifies the general algorithm for the process so that it can be automated in a data mining technique In summary, the paper finds ways for the investor to reduce the short term risk of investing in the share market.

Critical Review of Data Mining Techniques for Gene Expression Analysis

Classification of gene expression data has been exploded in the recent years. This can aid in the development of efficient methodology in the field of bio-informatics to be used for tumours diagnosis and treatment. Data mining is an effective technique being used in this field. One of the most difficulties facing this technology is the inappropriate classification methods that examine complex structure of gene expression data. In this paper, we give a brief introduction of gene expression data with experiment and we have made a critical review of major techniques being applied in the field of gene expression data with help of data mining. It can be seen that researchers have developed various techniques for gene data classification. In addition, they may differ from one to another whereas results are still showing the need for enhancement in this field. Some of these techniques are addressed in this paper in term of advantages and disadvantages. Accordingly, the deoxyribonucleic acid (DNA) is considered as the maestro of the tumour-derived factors. Analyzing changes on the gene expression may give rise for diagnosis enhancement of affected tissues in their early stages. For that reason, an ongoing research is addressing the problem of subspace clustering methodologies suitable for high dimensional datasets and verify of the new methodologies using appropriate datasets, particularly suitable for the analysis of gene expression data. In this context, researchers have identified various limitations of these methods particularly in the areas of information integration systems, text-mining and bio-informatics.

Fuzzy based binary feature profiling for modus operandi analysis

It is a well-known fact that some criminals follow perpetual methods of operations, known as modus operandi (MO) which is commonly used to describe the habits in committing something especially in the context of criminal investigations. These modus operandi are then used in relating criminals to other crimes where the suspect has not yet been recognized. This paper presents a method which is focused on identifying the perpetual modus operandi of criminals by analyzing their previous convictions. The method involves in generating a feature matrix for a particular suspect based on the flow of events. Then, based on the feature matrix, two representative modus operandi are generated: complete modus operandi and dynamic modus operandi. These two representative modus operandi will be compared with the flow of events of the crime in order to investigate and relate a particular criminal. This comparison uses several operations to generate two other outputs: completeness probability and deviation probability. These two outcomes are used as inputs to a fuzzy inference system to generate a score value which is used in providing a measurement for the similarity between the suspect and the crime at hand. The method was evaluated using actual crime data and four other open data sets. Then ROC analysis was performed to justify the validity and the generalizability of the proposed method. In addition, comparison with five other classification algorithms showed that the proposed method performs competitively with other related methods.

A comparison of clustering algorithms in categorizing economic events based on the behavior of exchange rates

Cluster analysis is used to identify dissimilar subgroups of objects out of a set of objects based on a combination of rules. In the light of cluster analysis, it is possible to treat dissimilar individuals in an appropriate manner by taking their dissimilarity into consideration. This will be resulted in enhancing the accuracy and efficiency of estimation and prediction models. This study aims to evaluate the performance of different partitioning methods namely, k-means, k-medoids (PAM) and fuzzy and hierarchical methods namely, agglomerative nesting and divisive analysis in grouping the economic events affecting the foreign exchange market. Cluster analysis performed on economic indicators data set depicts the structure of clusters resulted from all algorithms are the same except the single linkage of agglomerative nesting. Poor quality of the clustering structure formed by the single linkage method is confirmed by the lower value of average silhouette width. Comparatively high value of agglomerative coefficient associated with the ward’s method reveals the better performance of clustering compared to other linkages. Economic indicators under study are found to be clustered in three groups as performing high, moderate and low impact on the movements of exchange rates. High impact of economic indicators on the exchange rates is reflected by the high volatility at release time and shorter prevailing time of the impact after the release. Keywords-clustering algorithms; partitioning methods; hierachichal methods; foreign exchange; volatility; economic indicators

Analysis of high dimensionality yeast gene expression data using data mining

Data Mining is the process of discovering interesting knowledge from large amounts of data stored either in databases, data warehouses, or other information repositories. From biological studies, the Yeast Proteome Database (YPD) is a model for the organization and presentation of genome-wide functional data. Accordingly, a yeast gene expression which is a unicellular DNA is selected which contains 6103 genes and the database combined with a number of related dataset to create a general dataset. DNA-binding transcriptional regulators interpret the genome’s regulatory code by binding to specific sequences to induce or repress gene expression. The gene products including RNA and protein are responsible for the development and functioning of all living membranes by 2 steps process, transcription and translation. Various transcription factors control gene transcription by binding to the promoter regions. Translation is the production of proteins from mRNA produced in transcription. In this study, out of the 169 transcription factors known to access yeast, we are considering those thought to be involved in the response of Hydrogen Peroxide (H2O2). They are 22 transcription factors. Each one is partitioned to 3 parts: TF with No H2O2, TF with Low H2O2 and TF with High H2O2. The aim of this paper was to enhance the effectiveness of the integration of hydrogen peroxide response data related to yeast gene expression data to obtain a protein response process model and to label a set of important genes related to this approach.