Page 186 - AI Computer 10

Let us understand the association model with the help of a simple scenario. Suppose, you have a retail dataset
containing transactions from various customers and you want to predict whether a customer will buy a specific
item based on their previous shopping behaviour. In such a situation, you can use the association method.
Step 1: Finding patterns in purchase behaviour by writing a rule like:

{Bread}  {Butter}
Step 2: Create binary features based
on the presence of valuable
association rules. For example,
"Buys_Butter_if_Bread": This
feature is set to 1 if a customer
has a history of buying bread and
butter, and 0 otherwise.

Step 3: Use the newly created features
alongside other existing features
to train a model to predict
whether a customer will buy a new product.

Step 4: Train the model and evaluate its predictive performance.
Some other applications of Association model are:
u Cross-selling in banking, where a customer who took a vehicle loan may be interested in buying an insurance
plan too.
u Personalized Marketing and Targeted ads, where a person buying camping gear may be interested in
purchasing clothes and footwear suitable for camping.
u Food Pairing and Recipe suggestions, where a person buying pasta may also need to buy sauces and
seasoning.

Pop Quiz Quiz
Pop
Classify the given scenarios as Cluster based model or Association based model.

1. A website wants to segment users into different groups based on their browsing behaviour, such as
Sports Enthusiasts, Food Lovers, and Tech Savvy __________________________________________

2. An e-commerce company wants to understand which products are commonly bought together to offer
bundled discounts. ________________________________________________________________________

3. A health insurance company wants to group policyholders into categories based on risk factors like
age, medical history, and lifestyle for better coverage plans. ___________________________________

4. A marketing team wants to identify frequently co-occurring keywords or hashtags in tweets to
identify popular trends. ____________________________________________________________________
5. An online retailer wants to group products based on customer preferences to display the most
relevant ones on the homepage. ______________________________________________________________

TYPES OF DEEP LEARNING MODELS

Deep Learning enables software to autonomously learn how to perform tasks using large volume of data. In Deep
Learning, machines are trained on extensive datasets, allowing them to self-optimise based on the information.
These systems can develop their own algorithms, showcasing a level of intelligence, similar to humans.

52
52