View Forum Topic

Forums
CS522
HW2-Part1

Author	Message
aaronyen Posts: 29	Posted 09:22 Feb 01, 2012 \| Hi Dr. Sun, For the second file of part 1, which contains the step-by-step description of applying the Apriori algorithm, do you want us to display how we get the strong association rules for all frequent itemsets or is one example suffice? Thank you. Last edited by aaronyen at 09:23 Feb 01, 2012.
cysun Posts: 2935	Posted 20:39 Feb 01, 2012 \| I want to see L1, C2, L2, C3, ..., and then every association rule generated from the frequent itemsets and their confidence, so yes, basically i want to see everything.
aaronyen Posts: 29	Posted 20:50 Feb 08, 2012 \| Hi Dr. Sun, In the example you gave us, A -> B, SupportCount, Confidence 48 170 -> 38 39, 1193, 0.7662170841361593 Does the above example give the exact SupportCount and Confidence of that particular association for retail.txt? I am asking because my program gives the correct output for my testing data. I am also getting 48 170 -> 38 39 for retail.txt, but the SupportCount is different while the Confidence is still 0.766. Lastly, is it okay if my output has brackets around the sets(eg. [1,2] -> [3])? Thank you very much.
cysun Posts: 2935	Posted 07:39 Feb 09, 2012 \| aaronyen wrote: Hi Dr. Sun, In the example you gave us, A -> B, SupportCount, Confidence 48 170 -> 38 39, 1193, 0.7662170841361593 Does the above example give the exact SupportCount and Confidence of that particular association for retail.txt? I am asking because my program gives the correct output for my testing data. I am also getting 48 170 -> 38 39 for retail.txt, but the SupportCount is different while the Confidence is still 0.766. Lastly, is it okay if my output has brackets around the sets(eg. [1,2] -> [3])? Thank you very much. It probably is, though I'm not 100% sure. It should be easy to write a simple program to verify the support count. The output format doesn't matter much as long as it contains all the information, so yes, feel free to add brackets.
cysun Posts: 2935	Posted 16:01 Feb 09, 2012 \| I checked and the support count is indeed 1193. What number did you get?
aaronyen Posts: 29	Posted 16:29 Feb 09, 2012 \| cysun wrote: I checked and the support count is indeed 1193. What number did you get? Hi Dr. Sun, For the count of {38,39,48,170} I have 1172 and for {48, 170} -> {38,39} I have 1531, which is around 0.766 confidence. My retail.txt has total of 88162 transactions, with the last one = {32, 39, 242, 1393}. Is that right? Thank you.
cysun Posts: 2935	Posted 17:17 Feb 09, 2012 \| aaronyen wrote: cysun wrote: I checked and the support count is indeed 1193. What number did you get? Hi Dr. Sun, For the count of {38,39,48,170} I have 1172 and for {48, 170} -> {38,39} I have 1531, which is around 0.766 confidence. My retail.txt has total of 88162 transactions, with the last one = {32, 39, 242, 1393}. Is that right? Thank you. Why do you have different counts for {38,39,48,170} and {48, 170} -> {38,39}?
aaronyen Posts: 29	Posted 17:54 Feb 09, 2012 \| cysun wrote: aaronyen wrote: cysun wrote: I checked and the support count is indeed 1193. What number did you get? Hi Dr. Sun, For the count of {38,39,48,170} I have 1172 and for {48, 170} -> {38,39} I have 1531, which is around 0.766 confidence. My retail.txt has total of 88162 transactions, with the last one = {32, 39, 242, 1393}. Is that right? Thank you. Why do you have different counts for {38,39,48,170} and {48, 170} -> {38,39}? Hi Dr. Sun, I've fixed it. Thank you. Last edited by aaronyen at 18:08 Feb 09, 2012.