what is ASsociation Rule mining?

Association Rule Mining (ARM) is another widely used machine learning technique. It falls into the category of unsupervised learning, where labels are not provided, and the model trains autonomously without human intervention. The primary goal of ARM is to analyze the data, produce patterns, and check for their associations/correlations. It involves the process of finding the frequently brought items(frequent item sets) and generating laws that establish relationships among different items(association rules). Some of the common applications of ARM include:

• Market Basket Analysis

• Text Mining

• Social Network Analysis

• Pricing Optimization

What ARE ruleS? 

Rules in ARM are defined as a set of relationships or associations between different itemsets(variables) present in the dataset. The rules are generally expressed in the form of antecedent and consequent expressing the associations between them. The rules aim to capture patterns and associations within the data. 

For example, 

In this case, the rule can be framed in this way: 

If the weather is warm and the origin state is Florida, then there is a high likelihood that flights will be delayed. 

This can also be written as {Warm, Florida} --> {Delay}

where {Warm, Florida} is antecedent and {Delay} is consequent.

In this scenario, if the support of bananas does not meet the minimum threshold, Apriori prunes it. The algorithm then proceeds to create sets with other items that have surpassed the minimum support threshold.

How ARM is used in this project: 

The utilization of ARM in this project is crucial for examining patterns related to flight delays resulting from adverse weather conditions. Specifically, ARM is employed to identify flights that exhibit delays and to discern the climatic conditions under which these delays are more likely to occur. The analysis is specifically conducted on the origin states and their corresponding weather delays, given the substantially higher proportion of weather-related delays at the point of departure compared to the destination. The project involves the discovery and visualization of distinct sets of rules that highlight the significant relationships between specific weather conditions in origin states and the occurrence of flight delays.