Silmi Fauziati

Comparative study on data mining classification methods for cervical cancer prediction using pap smear results

The number of woman with cervical cancer in Indonesia is getting higher. Indonesia becomes the country with the highest number of women with cervical cancer in the world. Cervical cancer became the highest cause of cancer deaths in women globally. There has been a lot of research using data mining techniques with variety of different data mining models that can be used for analyzing cervical cancer. In this research, data that be used were obtained from the medical records of the Pap smear test results. There are 38 symptoms and 7 classes. Naïve Bayes, Support Vector Machines (SVM), and Random Forest Tree was used to evaluate the performance of the classifier. The performance matric that used in this study are accuracy, recall, precision, and ROC curve. Based on the performance matric, Random Forest Tree is the best classifier among other classifiers to classify Pap smear results.

Decision support system for stock trading using multiple indicators decision tree

Decision support system using decision tree classification can be used for stock trading technical analysis. Technical analysis is a stock analysis method which solely based on interpreting stocks price chart movement or trend. Historical stock prices and volume are used as variable input. The system is built based on financial market technical analysis indicators (Exponential Moving Average, Moving Average Convergence Divergence, Relative Strength Index, Money Flow Index, and parabolic Stop and Reverse). The proposed method is arrange indicators set into decision tree based on stock trading rules and it create buy, hold, and sell classes which represented decisions in trading. Decision classes then are analyzed for their profitability, geometric mean return, and cumulative wealth index. Furthermore sensitivity analysis is added into profitability analysis to obtain more positive value trading in decision making. The research purpose is to enhance decision making in technical stock trading. Compared to the single indicator decision tree multiple indicators offers 20% enhancement in decision making.

A comparative study on fuzzy Mamdani-Sugeno-Tsukamoto for the childhood tuberculosis diagnosis

World Health Organization (WHO) estimated that approximately 80 thousand children died every year in view of Childhood Tuberculosis. The disease needs an appropriate treatment considering the difficulties in establishing a diagnosis in pediatric patients. The incapability of children to produce sputum becomes one of the difficulties. Sputum is used to diagnose a person suffering from tuberculosis, based on Mycobacterium tuberculosis in sputum. In this paper, Mamdani, Tsukamoto and Sugeno-types Fuzzy Inference System are applied to assist the tuberculosis diagnosis. The different technique in these three methods is aimed to determine the most appropriate method for such diagnosis. The results show that, of the three types of Fuzzy Inference System, the best model is Sugeno model. Sugeno-type FIS has a better accuracy compared to both Mamdani and Tsukamoto ones at 93%, equivalent to a fault diagnosis in 13 of 180 patients. Here, Mamdani-type FIS is provided the diagnostic accuracy of 89%, equivalent to the ...

I forex trend prediction technique using multiple indicators and multiple pairs correlations DSS: A software design

Technical analysis is a method to forecast market price quickly. It has a paradigm which stated that history repeats itself. According to technical analysis theory, historical data is an important variable to predict the future. The previous variables that used to calculate are open, high, low, and close price. There are many technical ways to analyze and predict trends using one kind of pair. This paper proposes approach that provides predictions and analyses by comparing multiple pairs to give strength measurement using their correlations. The Decision Support Systems correlate the calculation result from Moving Average, Relative Strength Index, Parabolic Stop and Reverse, and William %Range. It resulted the trend strength that can be used to ensure the power and validity of a trend being, so the trend prediction is more accurate. In the research period of 78 weekdays the signal appeared 7 times, in which 4 of 7 signals are valid. All of false signal occurred when the predicted trends strength is less than 50%.

ARIMA implementation to predict the amount of antiseptic medicine usage in veterinary hospital

The medicine is a tool for patients either human or animal to treat or prevent disease. One of the medicines which often used to help prevent infections is an antiseptic medicine. Povidone-iodine is an antiseptic and disinfectant medicine that is extensively used by hospital and other health care. A health care institution such as pharmacies or hospitals must ensure the availability of medicines for patients. These things make the pharmaceutical dealing with the uncertainty of the medicine needed. Veterinary hospital as a health service provider has several challenges. One of the challenges is it must ensure the availability of the medicine needed at all time. The predictive capability of the amount of medicine usage can assist for managing the availability of medicines in veterinary hospitals. This study analyses and presents a forecasting model using ARIMA method to predict an antiseptic povidone-iodine usage at Prof. Soeparwi Veterinary Hospital. ARIMA method was selected because the method can be used in fitting forecasting models when there is a non-stationary time series and the method has capability to correct for local trend in data. Results of this study is ARIMA (1,0,1) was selected as a suitable model to represent historical dataset.

Management information systems development for veterinary hospital patient registration using first in first out algorithm

Registration system is one of the most important elements in an organization or institution that involves the presence of customers and one by one services. Nowadays, with the development of technology, services process in an institution become more effective and efficient. One of the institutions that require the development of this technology, in the form of hospital information management system, is the veterinary hospital. The conventional system has many shortcomings include allowing an error in writing the patient data also the patient registration recaps is less effective and more time consuming. Therefore, a prototype of RSH management information system was made as an ilustration in the design of hospital information management system. The focus of this study is the patient registration system which use First In First Out (FIFO) algorithms where the patient who came first to the hospital is the one who enrolled first.

Pattern of accesibility level of health facilities in yogyakarta

Regional development programs in the regional autonomy policy require the smallest area unit information. One of those programs is the construction of health facilities. The process of health development in the future requires the pattern of distribution of health facilities that already exists today. Distribution of health facilities in the province of Yogyakarta is still concentrated in urban areas, and the number of health facilities is not comparable with the area of each district. Therefore spatial data analysis using Exploratory Spatial Data Analysis (ESDA) is needed in order to see the pattern of accessibility level of existing health facilities. Spatial analysis considers the spatial correlation among the regions as the object of research. The existence of health facilities in a village will influence the accessibility to health facilities in another village. This paper uses health facilities index to find spatial clusters of accessibility level. New method based on administrative boundaries is added to this analysis as weights matrix to be compared with existing methods in GeoDa tools. The experiments show the use of weights matrix by administrative boundaries gives the best result. This weighing has the highest Moran’s I, so it is more suitable weighing method for this analysis. Furthermore, the spatial clustering of the easiest access to health facilities occurs in the urban areas and some area around them, such as in Yogyakarta City and Sleman Regency near Yogyakarta City, also in central government area.

State of charge estimation of Lithium Polymer battery using ANFIS and IT2FLS

in this research, the estimation method using IT2FLS (Interval Type 2 Fuzzy Logic System) and ANFIS (Adaptive Neuro-Fuzzy Inference System) as a base to build the membership functions and the rule base is constructed. The differences area of uncertainty is used to determine a model of type 2 fuzzy systems based on the smallest RMSE value. This study uses two methods of type-reducer, namely Enhanced Iterative Algorithm with Stop Condition (EIASC) and Enhanced Opposite Direction Search (EODS) to determine the most appropriate capacity estimation of the battery. Two types of datasets are used to determine the method performance indicated by MSE, RMSE and MAE. Based on the tests performed in three methods: T1FLS, IT2FLS EIASC, and IT2FLS EODS, it has been found that IT2FLS produces the smallest RMSE value with the RMSE value of 3.3% for static discharge dataset and 5.9% for pulse variation dataset.

Development of semi-supervised named entity recognition to discover new tourism places

Tourism information needs are increasing in line with tourism that has been a primary need for some people. This has an impact on the growth of the tourism information provider. The amount of available information sometimes makes tourist confuse to the information that they needed. Currently, the search systems only rely on indexing web pages so that the information obtained by the tourist is still unfavorable because it only shows a web page with keywords that exist on the article. A support system to recognize tourism places on the web pages is required to produce better information presentation. In this study, the recognition system based on Yet Another Two Stage Idea (YATSI) Semi-Supervised Learning with the Naïve Bayes classifier is used to address the problem. Results obtained by classifying candidate entities on a hundred web pages demonstrate 74% precision with 70% recall.

Combat aircraft effectiveness assessment using hybrid multi-criteria decision making methodology

Selection of military defense equipment, especially combat aircraft appropriately, effectively and efficiently affects to the readiness of the Indonesian Air Force in upholding the countrys sovereignty in the air. Based on the problems, the military requires an application that can support decision-making for the electoral system Combat Aircraft. This paper presents a decision support system using MCDM method with a combination of methods Analytic Hierarchy Process (AHP) and Method Techniques For Preference Order By Similarity (TOPSIS).The methods use an experimental design technique to assign weights attributes and then the operating methods for building models of decision making. As an illustration of this model using six examples of anaircraft type that are still in operation this era with 6 kinds of criteria which determine in air to air dogfight. By using the MCDM method will generate preference value to rank each alternative type of aircraft.