Implementation of Adaptive Neuro-Fuzzy Inference System (ANFIS) Algorithm for Customer Credit Prediction.pptx

sebuah file persentasi mengenai penelitian

1. Implementation of Adaptive Neuro-Fuzzy Inference System (ANFIS) Algorithm for Customer Credit Prediction Edy victor haryanto S mikha dayan sinaga nita sari sembiring noprita Elisabeth sianturi

2. introduction Nowadays, credit policies in the sales process are one of the supplier's most powerful tools for controlling demand, increasing sales, and promoting commodities. Any supply chain member wishing to purchase obtains full or partial trade credit based on strong decision-making rights. The problem that often occurs is the large number of bad loans that can disrupt the economic cycle. This problem has also resulted in many finance companies experiencing difficulties in processing customer loans. This problem can be prevented by applying artificial intelligence to predict which customers will take credit Predictions will be made using the ANFIS method.

3. method The research conducted has differences with previous research. This study takes sample data from finance companies that usually carry out the process of financing automotive product loans. The assessment criteria used are also different from those in previous studies. This research will produce a model using the anfis algorithm that can be used to predict whether the prospective customer will be able to pay credit on time or not. This research is also expected to be able to help with bad credit problems that often occur in finance companies in Indonesia. Method use in this research is Adaptive Neuro Fuzzy Interference System (ANFIS ), ANFIS is an algorithm that combines a fuzzy system with an artificial neural network system

4. result D ata R equirement Data obtained from one of the financing companies in the city of Medan. The data used is 200 data consisting of several criteria as follows and table 1. 1. Character (Government Employees (GE), Private Employees (PE), Entrepreneur (EN)). 2. Capacity (High (HI), Average (AV), Low (LO)). 3. Capital (High (HI), Average (AV), Low (LO)). 4. Condition (Good, Poor).

5. Data Transformation The data obtained is then converted into a form that can be processed by the ANFIS model. The data is changed based on several assessment criteria. Based on the assessment criteria, the data that has been changed can be seen in the following table 2

6. The basic structure of the ANFIS network can be seen in Figure 1 below.

7. Data testing with the ANFIS method was carried out using Matlab software. The test results between the results from the previous data and the results from the ANFIS method can be seen in Figure below

8. Testing the ANFIS method with Matlab also produces ANFIS networks which can be seen in Figure below. From the results of data testing using the ANFIS method, 81 rules were formed that can be used to predict customer data

9. The results of the comparison between the previous data calculations with the ANFIS method can be seen in the following table below: Based on table 3 above, we can see that the results obtained from the results of the analysis using the ANFIS method were 39 data that were not suitable from the 200 data analyzed or 19.5% data that were not suitable and 80.5% data that were suitable

10. The results obtained are considered HI enough to be used as the success rate of the model applied to the ANFIS method to be able to predict future customer data

11. conclusion Based on the research that has been done, it is known that many finance companies experience bad credit problems from their customers. The problems experienced by the company can be solved by predicting whether the prospective employee can make payments on time or not. The research carried out gave results of 39 data from the 200 data studied or 80.5% of the data according to the predictions obtained from the results of the analysis of the ANFIS method with previous data. This indicates that the ANFIS method can be used to predict customer credit payments. These results can also be used as a reference or input for companies in providing credit approval for their prospective customers. This can reduce the burden on the finance company due to the large number of bad loans from customers and can make the company a financially healthy company. Future research will try to apply other machine learning methods to see which method is the best to use in predicting the credit of prospective customers at finance companie

12. Thank You