Machine learning algorithm commonly used association rules
Association rules algorithm is a rule-based machine learning algorithm, the algorithm can be found in relation to each other in big data.
Its purpose is to: use some metrics to distinguish between strong rules exist in the database. That association rule mining is a knowledge discovery, rather than the forecast, it is part of unsupervised machine learning methods.
Algorithm name | Algorithm Description |
---|---|
Apriori | Association rules most commonly used is the most classical algorithm for mining frequent item sets The core idea is to produce candidates for their support through the connection and then generate frequent item set by pruning |
FP-Tree | The defects inherent in the multiple scan the transaction data set Apriori algorithm, the proposed method does not generate candidate frequent item sets. Apriori and FP-Tree is an algorithm to find frequent item sets |
Eclat algorithm | Eclat algorithm is a depth-first algorithm, vertical data representation, based on the use of equivalence relations prefix search space is divided into smaller sub-space based on theoretical concepts |
Gray Correlation Method | An analytical method for analyzing and determining the degree of influence between the factors, or a number of sub factors (subsequence) of the main factors (parent sequences) and the degree of contribution made |
Recommended Reference article:
Apriori algorithm using correlation analysis
https://blog.csdn.net/LaputaFallen/article/details/78609881
Association Rules and FP-Tree
https://blog.csdn.net/lpsl1882/article/details/52593855
Frequent item set mining algorithms --Eclat algorithm
https://blog.csdn.net/my_learning_road/article/details/79728389
Gray Relational Analysis (GRA) of Theory and Application (matlab and python)
https://blog.csdn.net/starter_____/article/details/82085040