What is Artificial Intelligence
• Artificial intelligence is a machine’s ability to perform some
cognitive functions we usually associate with human minds.
• Artificial intelligence (AI) Session
can be 8
defined as applying
advanced analysis and logic-based techniques, including
machine learning (ML), to interpret events, support and
automate decisions, and take actions.
Ref: https://www.gartner.com/en/topics/artificial-intelligence
https://www.mckinsey.com/featured-insights/mckinsey-explainers/what-is-ai#/
The Evolution of Artificial Intelligence
Ref: https://www.calls9.com/blogs/the-history-of-ai-a-timeline-from-1940-to-2023
Machine Learning
Machine Learning algorithms detect patterns in large data sets and learn
to make predictions by processing data
Regression
Supervised Learning
Classification
Machine
learning
Clustering
Unsupervised Learning
Anomaly Detection
Dimensionality
Reduction
Supervised Learning
Supervised learning learns from data labelled into right answers
Regression predicts a number from Classification predicts categories from a
infinitely many possible outputs finite limited set of outputs
Ref: https://applicationbooy.weebly.com/blog/excel-regression-analysis-non-numeric-data
https://noeliagorod.com/2019/05/21/machine-learning-for-everyone-in-simple-words-with-real-world-examples-yes-again/
Unsupervised Learning
Unsupervised learning finds something interesting in unlabeled data. Data
comes with only inputs X but not outputs Y
Clustering groups data Anomaly detection identifies Dimensionality reduction reduces the
according to observed data not fitting the groups number of variables in a dataset while
patterns preserving its most important features.
Eg: Google news search Eg: Fraud detection by Eg: For biological research where the
grouping related items identification of unexpected quantity of genetic variables outweigh
together behavior the number of observations.
Deep Learning
• Deep Learning, usually neural networks inspired by the ways neurons
interact in the human brain to ingest data and process it through multiple
iterations that learn increasingly complex features of the data and make
increasingly sophisticated predictions
• Neural networks can help computers make intelligent decisions with limited
human assistance. This is because they can learn and model the
relationships between input and output data that are nonlinear and complex.
• Neural networks can comprehend unstructured data and make general
observations without explicit training. For instance, they can recognize
that two different input sentences have a similar meaning
Ref: https://aws.amazon.com/what-is/neural-network/
Some Applications of Neural Networks
Computer vision Speech recognition Natural language processing Recommendation engines
• Computer vision is the • Neural networks can • Natural language processing • Neural networks can track
ability of computers to analyze human speech (NLP) is the ability to user activity to develop
extract information despite varying speech process natural, human- personalized
and insights from patterns, pitch, tone, created text. Neural recommendations.
images and videos. language, and accent. networks help computers
• They can also analyze all
gather insights and meaning
user behavior and discover
• With neural networks, • Virtual assistants like from text data and
new products or services
computers can Amazon Alexa and documents.
that interest a specific user.
distinguish and automatic transcription
recognize images software use speech • NLP has several use cases • For example, it can help
similar to humans. recognition to do tasks. such as automated virtual brands convert social
agents and chatbots. media posts into sales.
• Visual recognition in • Converting clinical Consumers don't have to
self-driving cars so conversations into • Business intelligence hunt through online
they can recognize documentation in real time analysis of long-form catalogs to find a specific
road signs and other also utilizes speech documents like emails and product from a social
road users recognition forms also utilizes NLP media image.
Ref: https://aws.amazon.com/what-is/neural-network/
Architecture of Neural Networks
The four most common types of neural network layers are-
•Fully connected layer- In a fully connected layer, each neuron
is connected to every neuron in the previous and next layers,
allowing for complex relationships between input and output.
•Convolutional layer- Convolutional layers use filters to
extract spatial features from input data, enabling effective
pattern recognition in images or sequential data.
•Deconvolutional layer- Deconvolutional layers, also known as
transposed convolutional layers, reverse the operation of a
convolutional layer by upsampling the input to reconstruct the
original image or data.
•Recurrent layer- Recurrent layers incorporate memory units
to process sequential data by capturing dependencies and
interactions among previous inputs, making them suitable for
tasks involving time series or natural language processing.
Generative AI
• Generative AI is a branch of deep learning that uses exceptionally
large neural networks called large language models (LLM) that can
learn specially abstract patterns. LLM can be applied to interpret and
create text, video images and data
• Generative AI offers enormous productivity benefits for individuals
and organizations, and while it also presents very real challenges and
risks, businesses are forging ahead, exploring how the technology can
improve their internal workflows and enrich their products and
services
https://www.ibm.com/think/topics/generative-ai
Types of Gen AI models
Ref: https://blog.gopenai.com/an-overview-of-7-types-of-generative-ai-models-c1581f0f57e6
Harrah’s Entertainment
What are the options that Harrah’s may have leveraged for
its database marketing & performance analysis?
• Descriptive Analytics: (What has happened?) (Data aggregation,
Summary)
• Predictive Analytics : (What might happen?) (Regression, Neural
Network)
• Prescriptive Analytics : (What should we do?) (Optimization,
Recommendation)
Opportunity based customer segmentation
Segmentation is a case of categorization or clustering
Harrah’s is aiming for a customer centric approach instead of direct marketing
New Business Program
• Designed to convert total Gold members to loyal customers
• Uses predicted customer worth (theoretical wins) to make competitive
offers
Loyalty Program – Frequency Upside
• Designed to identify and convert customers who were giving Harrah’s
only a small share of their spending
• Purpose was to make them shift from competitor to Harrah’s
• Profitability calculated by comparing theoretical incremental wins to
cost of the program
Loyalty Program – Budget Upside
• Designed to identify customers with budget upside
• Allocating small share of gaming budget to Harrah’s
• The first casino visited received maximum budget
Retention Program
• Designed to reinvigorate customers who had broken their historical
visitation pattern or had demonstrated other signs of attrition
• Tested variety of offers with retained customers to determine how
much to reinvest
• Based on historical patterns ( Regression??) they expected to see them
in December but did not..
Total Rewards Program
• Program designed to facilitate cross market visitation patterns
• To address estimated loss of $100 million generated by Harrah’s
customers in Las Vegas alone
• Hinged on making it easier for earn and redeem rewards seamlessly at
any Harrah’s property across country
• Designed an integrated IT systems to manage the program
References
• https://www.coursera.org/learn/machine-learning
