An Integrated Methodology of Artificial Neural Network and Rough Set Theory for Analyzing IVF Data

This paper illustrates the process of applying data mining for finding success rate of In-Vitro Fertilization treatment. The data set used in the experiments contains information recorded during the IVF treatment. In the research paper defined the supportive information to the medical practioner for knowing the success rate of patient before starting the Artificial Insemination (11). In the IVF, the doctors and patients may don´t know the way of predicting the success rate of the treatment (9). The success rate may help the patients to be getting ready for the treatment physically and psychologically (8). In data mining has many tools for data reduction and prediction (5). Rough Set Theory (RST) used for the data cleaning and reduction (13). It presents the influential parameters of the IVF treatment as an output. The Artificial Neural Network (ANN) gets the output of the RST as an input and built a network for the input and produce desired output. So the processes checked the result of patient and compare the desired and actual output (6). This experiment is a way of study which is related to the representativeness of the sample and irrelevant features. Out of around 250 million individuals estimated to be attempting parenthood at any given time, 13 to 19 million couples are likely to be infertile. So the couples prefer the IVF treatment compared with other methods of treatment (5). In India the board of medical council announced the duration of infertility. If a woman was not conceived after his marriage within 6 months they caused infertility [14]. So they must start the initial fertility treatment. Most of them prefer the In-Vitro fertilization compare with other fertility treatments (3). A survey of fertility treatment showed 1 in 20 of all pregnancies conceived by the IVF treatment. But the patients suffer from the negative imagination and they don´t know the success level of the treatment (1). It is very essential to analyze the data set and reduce or clea- the unwanted data that increases the prediction accuracy in a proper manner (15). The parameters with high impact factor can be selected by applying the proper reduct algorithm which removes the parameters that has a lesser role in determining the success rate of particular patients and help the Gynecologists to recommend them for specific treatment of IVF, IUI or ICSI (1).