DataMining

Assignment-1

How to compile?

./compile.sh

How to run?

./2017CS10341.sh input X -apriori output ./2017CS10341.sh output X -fptree output
./2017CS10341.sh input -plot

List of files submitted

1. fptree.cpp : The source code of FP-Tree Algorithm
2. apriori.cpp : The source code of Apriori Algorithm
3. plot.py : Script to generate the running time plots
4. compile.sh : Commands to compile source code
5. 2017CS10341.sh : Executing the code

Observations

We observed that the runtime of FP-Tree algorithm at lower threshold values is very better than that of Apriori algorithm.
We also observed that FP-Tree scales well with the size of dataset where as Apriori doesnot scale well with the dataset size.
This is because of two main reasons:

1. Apriori algorithm requires multiple database scans(size of the largest frequent itemset) while FP-Tree algorithm requires only two database scans.
   
2. Apriori first generates the candidate sets and then prunes the sets(bigger memory requirement) while FP-tree encodes the sets in a tree structure.
   

Names: Kamalesh Neerasa 2017CS10341 Balleda Kiran Bharga 2019MCS2792