Browse by Author
Results for Family Name : "Rahim" AND Given Name/Initial : "Susilawati"
|
Pengkelasan keberkesanan perancang keluarga menggunakan pepohon keputusan
Item Type: Conference or Workshop Item
Editor:
Year: 00/00/2012
Abstract: Awareness of the importance of family planning practices for a family institution in Malaysia began around the 1960s. In order to plan family size it is important to determine the appropriate and effective method of family planning. Therefore, this study was conducted to develop a best classification model and identify the factors of effectiveness of such methods and also for women who do not use any family planning method, factors that influence the effectiveness of non-use of methods are also identified. Classification techniques in data mining were used to obtain patterns of effectiveness of the family planning methods used. The classification techniques used are result trees J48, ID3, Random Tree, REPTree and Bayes classifier Naive Bayes. Secondary data from the fourth set of the Malaysian Population and Family Survey (MPFS) obtained from the National Population and Family Development Board (LPPKN) were used in this study. To obtain the best classification model, 10-fold cross validation method, Weka and Mynda applications were used. The test results found that the J48 result tree was selected as the best classification model because this model provides the highest accuracy values for both data sets of 87.83% and 80.42% compared to other techniques. The study has produced a decision tree that can generate a set of tips to identify the effectiveness of family planning methods used and a set of tips to identify the effectiveness of those who do not use family planning methods. In conclusion, this study does not cover women who are using family planning methods and it is hoped that future studies will cover the effectiveness of the method for women who are using family planning methods and make pattern comparisons based on MPFS 1, MPFS 2, MPFS 3 and MPFS 4 data sets.
|
|
|
|
|
