Efficient and Effective Induction of First Order Decision Lists [chapter]

Mary Elaine Califf
2003 Lecture Notes in Computer Science  
We present BUFOIDL, a new bottom-up algorithm for learning first order decision lists. Although first order decision lists have potential as a representation for learning concepts that include exceptions, such as language constructs, previous systems suffered from limitations that we seek to overcome in BUFOIDL. We present experiments comparing BUFOIDL to previous work in the area, demonstrating the system's potential.
doi:10.1007/3-540-36468-4_2 fatcat:6p64puklqndwxg2kqtop4ki57q