Data Science Apriori Algorithm in Python

Introduction to Apriori Algorithm in Python

The Apriori algorithm is a classical algorithm in Data Mining that is used for mining frequent itemsets and association rule mining.

Watch this Data Science Full Course Video to learn more:

What is Association Rule Mining?

As mentioned before, the Apriori algorithm is used for association rule mining. Now, what is association rule mining? Association rule mining is a technique to identify frequent patterns and associations among a set of items.

For example, understanding customer buying habits. By finding correlations and associations between different items that customers place in their ‘shopping basket,’ recurring patterns can be derived.

Say, Joshua goes to buy a bottle of wine from the supermarket. He also grabs a couple of chips as well. The manager there analyses that, not only Joshua, people often tend to buy wine and chips together. After finding out the pattern, the manager starts to arrange these items together and notices an increase in sales.

This process of identifying an association between products/items is called association rule mining. To implement association rule mining, many algorithms have been developed. The Apriori algorithm is one of the most popular and arguably the most efficient algorithms among them. Let us discuss what an Apriori algorithm is.

What Is an Apriori Algorithm?

Apriori algorithm assumes that any subset of a frequent itemset must be frequent.

Say, a transaction containing {wine, chips, bread} also contains {wine, bread}. So, according to the principle of Apriori, if {wine, chips, bread} is frequent, then {wine, bread} must also be frequent.

How Does the Apriori Algorithm Work?

The key concept in the Apriori algorithm is that it assumes all subsets of a frequent itemset to be frequent. Similarly, for any infrequent itemset, all its supersets must also be infrequent.

Let us try and understand the workings of an Apriori algorithm with the help of a very famous business scenario, market basket analysis.

Here is a dataset consisting of six transactions in an hour. Each transaction is a combination of 0s and 1s, where 0 represents the absence of an item and 1 represents the presence of it.

We can find multiple rules in this scenario. For example, in a transaction of wine, chips, and bread, if wine and chips are bought, then customers also buy bread.

To select the interesting rules out of multiple possible rules from this small business scenario, we will be using the following measures:

• Support
• Confidence
• List
• Conviction

Support

Support of item x is nothing but the ratio of the number of transactions in which item x appears to the total number of transactions.

i.e.,

Support(wine) =

Support(wine) = = 0.66667

Confidence

Confidence (x => y) signifies the likelihood of the item y being purchased when item x is purchased. This method takes into account the popularity of item x.

i.e.,

Conf({wine, chips} => {bread}) =

Conf({wine, chips} => {bread})=  = 0.667

Lift

Lift (x => y) is nothing but the ‘interestingness’ or the likelihood of the item y being purchased when item x is sold. Unlike confidence (x => y), this method takes into account the popularity of the item y.

i.e.,

lift ({wine, chips} => {bread}) =

lift ({wine, chips} => {bread}) ==1

• Lift (x => y) = 1 means that there is no correlation within the itemset.
• Lift (x => y) > 1 means that there is a positive correlation within the itemset, i.e., products in the itemset, x and y, are more likely to be bought together.
• Lift (x => y) < 1 means that there is a negative correlation within the itemset, i.e., products in itemset, x and y, are unlikely to be bought together.

Conviction

Conviction of a rule can be defined as follows:

conv(x => y) =

i.e.,

conv({wine, chips} => {bread} ) = =  =1

Its value range is [0, +∞].

• Conv(x => y) = 1 means that x has no relation with y.
• Greater the conviction higher the interest in the rule.

Now that we know the methods to find out the interesting rules, let us go back to the example. Before we get started, let us fix the support threshold to 50 percent.

Step 1: Create a frequency table of all the items that occur in all transactions

Step 2: Find the significant items based on the support threshold

Support threshold = 3

Step 3: From the significant items, make possible pairs irrespective of the order

Step 4: Again, find the significant items based on the support threshold

Step 5: Now, make a set of three items that are bought together based on the significant items from Step 4

{Wine, Bread, Milk} is the only significant item set we have got from the given data. But in real-world scenarios, we would have dozens of items to build rules from. Then, we might have to make four/five-pair itemsets.

Watch Apriori Algorithm Tutorial

Hands-on: Apriori Algorithm in Python- Market Basket Analysis

Problem Statement

The manager of a retail store is trying to find out an association rule between six items, to figure out which items are more often bought together so that he can keep the items together in order to increase sales.

Dataset

Below is the transaction data from Day 1. This dataset contains 6 items and 22 transaction records.

Environment Setup:

Before we move forward, we need to install the ‘apyori’ package first.

pip install apyori

Market Basket Analysis Implementation within Python

With the help of the apyori package, we will be implementing the Apriori algorithm in order to help the manager in market basket analysis.

Step 1: Import the libraries

Step 2: Load the dataset

Step 3: Have a glance at the records

Step 4: Look at the shape

Step 5: Convert Pandas DataFrame into a list of lists

Step 6: Build the Apriori model

Step 7: Print out the number of rules

Step 8: Have a glance at the rule

The support value for the first rule is 0.5. This number is calculated by dividing the number of transactions containing ‘Milk,’ ‘Bread,’ and ‘Butter’ by the total number of transactions.

The confidence level for the rule is 0.846, which shows that out of all the transactions that contain both “Milk” and “Bread”, 84.6 % contain ‘Butter’ too.

The lift of 1.241 tells us that ‘Butter’ is 1.241 times more likely to be bought by the customers who buy both ‘Milk’ and ‘Butter’ compared to the default likelihood of sale of ‘Butter.’

Limitations of Apriori Algorithm

Despite being simple one, Apriori algorithms have some limitations including:

• Waste of time when it comes to handling a large number of candidates with frequent itemsets.
• The efficiency of this algorithm goes down when there is a large number of transactions going on through a limited memory capacity. 
• Requires high computation power and needs to scan the entire database. 

Improvements

The following are the ways to improve the efficiency of the algorithm:

• Use hashing techniques to reduce the number of database scans.
• Do not take the infrequent transaction further into consideration.
• If a purchase is frequent in one partition, it should be frequent in another partition. 
• Try to pick up random samples to improve the accuracy of your algorithm. 
• Use dynamic itemset counting to introduce new candidate itemsets while the scanning of the database is performed. 

Applications of Apriori Algorithm

Some of the popular applications of the algorithm are:

• Used in forest departments to understand the intensity and probability of forest fires. 
• Used by Google and other search engines for their auto-complete features. 
• The Healthcare department used such algorithms to analyze the patients’ database and predict which patients might develop blood pressure, diabetes, or other common disease. 
• Used to categorize students based on their specialties and performance to improve their academic performance. 
• E-commerce websites use it in their recommendation systems to provide a better user experience.

What Did We Learn?

In this tutorial, we have learned what association rule mining is, and what the Apriori algorithm is, and with the help of an Apriori algorithm example, we learned how the Apriori algorithm works. In the end, we have built an Apriori model in Python programming language on market basket analysis. Python Programming Course is one of the most demanding skills right now in the market.