Model Data Mining dan Rasio Altman untuk Prediksi Financial Distress Perusahaan
Financial distress is the stage or situation of a decline in the company's financial condition before a business failure occurs. This failure is related to liquidation, where the company fails to run its operations for profit. Financial distress prediction methods have been long developed in the field of finance using Multiple Discriminant Analyst (MDA). The most widely used MDA method is the Altman model ratio. This method is used as an early warning to the company's financial condition. This study aims to analyze the neuro fuzzy data mining algorithm with input data from the financial ratio of the Altman model to predict the company's financial distress. Research data are from the Indonesia Stock Exchange (IDX) website at www.idx.co.id. Data population between 2012-2017 LQ45 issuers are 80 companies. Preliminary data from the company's financial statements are calculated by Altman's ratio to get zscore values ??of three types of categories. These values ??are arranged in time series of four backward periods which will be used as input data for the neuro fuzzy algorithm ANFIS model. The next step is the formation of algorithms, training, testing, simulation and model evaluation using the GUI application from the Matlab program. The results of the model are predictive values ??that are compared with the results of the Altman ratio calculation of predicted years. The final results of the study show the predicted value of the ANFIS model of the Triangle membership function looks very optimal, consistent, and efficient. It was proven that the average predicted value of 9.95% was close to the actual value of the calculated year, while the results of the ANFIS model of the Trapezoid, Gauss and G-bell membership functions seemed less than optimal.
Keywords: Financial Distress, Neuro Fuzzy, Prediction, Altman Ratio.