KLASIFIKASI PERUSAHAAN DI INDONESIA DENGAN MENGGUNAKAN PROBABILISTIC NEURAL NETWORK (Studi Kasus: Perusahaan yang Terdaftar di Bursa Efek Indonesia Tahun 2016)

Adi Waridi Basyirudin Arifin (Departemen Statistika, Fakultas Sains dan Matematika, Universitas Diponegoro)
Hasbi Yasin (Departemen Statistika, Fakultas Sains dan Matematika, Universitas Diponegoro)
Budi Warsito (Departemen Statistika, Fakultas Sains dan Matematika, Universitas Diponegoro)

Jurnal Gaussian
Vol 6, No 4 (2017): Jurnal Gaussian

Classification of company performance can be judged by looking at it’s financial status, whether poor or good state. In order to classifying the financial status, annual financial report will be required. By learning financial status of company, it would be useful to evaluate the performance of the company itself from management cause, or as an investor, making strategy for investment to certain company would be easier. Classification of company performance can be achieved by some approach, either parametric or non-parametric. Neural Network is one of non-parametric method. One of the models in Artificial Neural Network is Probabilistic Neural Network (PNN). PNN consists of four layers, i.e. input layer, pattern layer, addition layer, and output layer. The distance function used is the euclid distance and each class share the same values as their weights. By using the holdout method as an evaluation in honesty, the results show that modeling the company performance with PNN model has a very high accuracy. This is confirmed by the level of accuracy of the data model built on the training data is 100%, while trial data also got 100% accuracy.            Keywords : Classification of Company Performance, PNN, Holdout.