/BPNN-Classifier

A Back-Propagation Neural Network (BPNN) Classifier to identify karst sinkholes from the LiDAR data.

Primary LanguagePython

BP Neural Network Classifier to Identify Karst Sinkholes from LiDAR Data

An extra credit assignment for Temple University CIS4526 Machine Learning course.

Project descriptions

Please see descriptions for details. descriptions

Instructions

Dependency: matplotlib

  1. Adjust the hyperparameter of BP network :
    • 10-hiden
    • net = BPNNet(num, 10, 1);
    • net.train(train_data, iterations=1000, N=0.01, M=0.1);
  2. Simply run 'pa4.py';
  3. Results are saved in an auto-generated file called "pred.csv". For values:
    • positive : the closer it is to 1, the larger chance there exists a sinkhole;
    • negative: there is no skinhole.

Training Performance

Training Performance