Introduction
Related work
Association rule
Minimum support
Literature review automate minimum support
The proposed model
Experiment and result
Data experiment
No | Dataset | Description |
---|---|---|
1 | Chess [55] | Chess game dataset |
2 | Mushrooms [55] | This data set contains descriptions of hypothetical samples belonging to 23 species of Agaricus and Lepiota gilled mushrooms (pp. 500–525). Each species is labeled as either definitely edible, definitely poisonous, or maybe edible but not recommended. This last category was merged with the toxic category. There is no easy criteria for determining the edibility of a mushroom, according to the Guide; no rule like “leaflets three, let it be” for Poisonous Oak and Ivy |
3 | Skin [55] | The skin dataset is created by sampling B,G,R values from face photos of people of various ages (young, middle, and old), races (white, black, and asian), and genders from the FERET and PAL databases at random |
4 | Accidents [55] | data on traffic accidents that has been anonymised |
5 | Connect [55] | This database contains all legal 8-ply connect-4 positions in which neither player has yet won and the following move is not forced |
6 | Retail [55] | Transactions with a Belgian retailer that has yet to be named |
7 | RecordLink [55] | Individual data such as first and last names, gender, date of birth, and postal code were acquired over the course of several years by iterative insertions |
8 | Test | Example transaction dataset |
No | Dataset | Number of transactions | Number of items | Average number of items on each transaction |
---|---|---|---|---|
1 | Chess [55] | 3196 | 75 | 37 |
2 | Mushrooms [55] | 8416 | 119 | 23 |
3 | Skin [55] | 245,057 | 11 | 4 |
4 | Accidents [55] | 340,183 | 468 | 33.8 |
5 | Connect [55] | 67,557 | 129 | 43 |
6 | Retail [55] | 88,162 | 16,470 | 10.30 |
7 | RecordLink [55] | 574,913 | 29 | 10 |
8 | Test | 3 | 5 | 3 |
Result
Dataset | Density rate (average occurrence of each item in all transactions) | Minimum support (density/number of transactions) | Number of rules with lift ratio > 1 |
---|---|---|---|
Chess | 1577.413 | 49.35% | 320,577,653 |
Mushrooms | 1626.445 | 19.32% | 19,011,494 |
Skin | 8165.67 | 33.36% | 2 |
Accidents | 36,992.1812 | 11.59% | – |
Connect | 22,519 | 33.33% | – |
Retail | 74.61 | 0.08% | 33,170 |
RecordLink | 213,691.1154 | 37.16% | 9892 |
Test | 1.8 | 60% | 4 |
Minimum Support Value | Runtime (MS) | |||||||
---|---|---|---|---|---|---|---|---|
Chess | Mushroom | Skin | Accidents | Connect | Retail | RecordLink | Test | |
100% | 11 | 12 | 0 | 0 | 0 | 78 | 0 | 0 |
90% | 29 | 16 | 0 | 1 | 4820 | 78 | 1 | 0 |
80% | 469 | 16 | 0 | 2 | 78 | 0 | 0 | |
70% | 8340 | 16 | 0 | 14 | 78 | 1 | 0 | |
60% | 109,935 | 20 | 0 | 58 | 120 | 8 | 0 | |
50% | 880,936,478 | 31 | 0 | 306 | 94 | 20 | 0 | |
40% | 94 | 0 | 3341 | 93 | 12 | 0 | ||
30% | 531 | 0 | 117 | 85 | 0 | |||
20% | 10,030 | 0 | 109 | 131 | 0 | |||
10% | 800,330 | 1 | 94 | 184 | 0 | |||
5% | 4 | 140 | 216 | 1 | ||||
0% | 1002 |
Minimum Support Value | Runtime (MS) | |||||||
---|---|---|---|---|---|---|---|---|
Chess | Mushroom | Skin | Accidents | Connect | Retail | RecordLink | Test | |
100% | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 0 |
90% | 4 | 0 | 0 | 9 | 546 | 0 | 0 | 0 |
80% | 367 | 0 | 0 | 22 | 45,541 | 0 | 0 | 0 |
70% | 6125 | 0 | 0 | 35 | 996,275 | 0 | 4 | 0 |
60% | 121,658 | 0 | 0 | 159 | 0 | 7 | 0 | |
50% | 1,431,849 | 3 | 0 | 816 | 0 | 7 | 0 | |
40% | 4 | 0 | 6297 | 0 | 7 | 0 | ||
30% | 73 | 0 | 64,794 | 0 | 44 | 0 | ||
20% | 37,757 | 0 | 1,058,700 | 0 | 77 | 0 | ||
10% | 1,213,180 | 0 | 0 | 168 | 0 | |||
5% | 0 | 0 | 175 | 0 | ||||
0% | 4 | 2683 | 0 |