Landslide hazard mapping of Penang island using probabilistic methods and logistic regression

This paper summarizes landslide hazard mapping of Penang Island, Malaysia using probabilistic methods and logistic regression. Probabilistic methods such as frequency ratio, statistical index, certainty factor and landslide susceptibility analysis, and logistic regression were applied to produce landslide hazard maps. Correlations were used to determine effects and contributions of landslide-causative factor to landslide. Factors that show low correlations or proven to negatively contribute to the final results were ejected. Final results show logistic regression as the best method to predict landslide occurrences in Penang Island with an accuracy of 80.05%, followed by frequency ratio with 79.68%, landslide susceptibility analysis with 79.6%, statistical index with 79.38% and certainty factor with 79.37%.