Coding the program for the demand of AutoClass (continued)

2007/04/11

1. Today, I have encoded the program for the AutoClass. In these programs, I adopted the score and the tf*idf as my input feature values for two programs, respectively. Then, I tested the generated files on the Auto-Class successfully. During the programming, I only used three documents for my testing data. Tomorrow, I will experiment on the whole documents to determine which feature value can achieve better performance.
2. In the evening, I wrote the assembly’s chapter 6 overview. Then, I uploaded it on the Google Document.