1. A case in Neural Network - A Nerual Network for Bankruptcy Prediction

2. Based on a paper –Published in Decision Support Systems, 1994 –By Rick Wilson and Ramesh Sharda ANN Architecture –Three-layer (input-hidden-output) MLP –Backpropagation (supervised) learning network Training data –Small set of well-known financial ratios –Data available on bankruptcy outcomes Moody’s industrial manual (between 1975 and 1982)

3. Application Design Specifics –Five Input Nodes X1: Working capital/total assets X2: Retained earnings/total assets X3: Earnings before interest and taxes/total assets X4: Market value of equity/total debt X5: Sales/total assets –Single Output Node: Final classification for each firm Bankruptcy (0) or Nonbankruptcy (1) Development Tool: NeuroShell Bankruptcy Prediction with ANN

4. Network Structure

5. Bankruptcy Prediction with ANN Training & Testing –Data Set: 129 firms –Training Set: 74 firms - 38 bankrupt, 36 not –Test Set (Hold out set) : 55 firms - 27 bankrupt firms, 28 non-bankrupt firms –Two Methods are compared: The neural network A conventional discriminant analysis program

6. Compare training performance of Neural Network with that of Discriminant Analysis

8. Performance Comparison with Test(hold out) data set 50/50 NN0.818 DA0.745 *NN: Neural Networks *DA: Discriminant Analysis

9. Results –The neural network correctly predicted (50/50): 81.8 percent accuracy (NN) 74.5 percent accuracy (DA) –Accuracy of about 80 percent is usually acceptable for this problem domain Bankruptcy Prediction with ANN