A Statistical Prediction of Likely Distress in Nigeria Banking Sector Using a Neural Network Approach

One of the most significant threats to the economy of a nation is the bankruptcy of its banks. This study evaluates the susceptibility of Nigerian banks to failure with a view to identifying ratios and financial data that are sensitive to solvency of the bank. Further, a predictive model is generated to guide all stakeholders in the industry. Thirty quoted banks that had published Annual Reports for the year preceding the consolidation i.e. year 2004 were selected. They were examined for distress using the Multilayer Perceptron Neural Network Analysis. The model was used to analyze further reforms by the Central Bank of Nigeria using published Annual Reports of twenty quoted banks for the year 2008 and 2011. The model can thus be used for future prediction of failure in the Nigerian banking system.

Morphometric Analysis of Tor tambroides by Stepwise Discriminant and Neural Network Analysis

The population structure of the Tor tambroides was investigated with morphometric data (i.e. morphormetric measurement and truss measurement). A morphometric analysis was conducted to compare specimens from three waterfalls: Sunanta, Nan Chong Fa and Wang Muang waterfalls at Khao Nan National Park, Nakhon Si Thammarat, Southern Thailand. The results of stepwise discriminant analysis on seven morphometric variables and 21 truss variables per individual were the same as from a neural network. Fish from three waterfalls were separated into three groups based on their morphometric measurements. The morphometric data shows that the nerual network model performed better than the stepwise discriminant analysis.