UNDERSTANDING

ARTIFICIAL
INTELLIGENCE(AI)
What is Artificial Intelligence (AI)?
• Artificial Intelligence (AI) is a branch of computer science focused on creating intelligent systems that can mimic
and simulate human intelligence processes and behaviors.
• AI encompasses a wide range of technologies and applications, including:
o Machine Learning: Algorithms that enable systems to learn from data and make predictions or decisions without being explicitly
programmed.
o Natural Language Processing: Technologies that allow systems to understand, interpret, and generate human language.
o Robotics: Development of intelligent robots that can perceive, reason, and act autonomously in physical environments.
• The goal of AI is to create systems capable of performing tasks that typically require human intelligence, such as:
o Decision-making and problem-solving
o Language understanding and generation
o Learning and adapting to new information and experiences
• AI systems are designed to exhibit human-like cognitive abilities, including:
o Reasoning and logical thinking
o Knowledge representation and acquisition
o Planning and strategizing
o Pattern recognition and generalization
• At its core, AI involves developing algorithms and models that can learn from data, identify patterns, and make
predictions or recommendations without being explicitly programmed for every scenario.
• The aim of AI is to create systems that can match or exceed human intelligence across a wide range of domains,
continuously improving their performance through learning and adaptation.
History of AI
• 1950: The Turing Test • 1984: The term "AI winter" is coined
• 1951: SNARC • 1985: Bayesian networks causal analysis
• 1952: Samuel Checkers-Playing Program • 1988: A Statistical Approach to Language Translation
• 1956: The term "artificial intelligence" is coined • 1989: CNNs applicability demonstrated
• 1956: The Logic Theorist • 1997: Deep Blue defeats Garry Kasparov
• 1957: The General Problem Solver • 2011: Watson defeats humans on Jeopardy!
• 1959: The term "machine learning" is coined • 2011: Apple releases Siri
• 1966: ELIZA and Shakey • 2012: Breakthrough in image recognition – Google's deep
neural network project
• 1973: The report "Artificial Intelligence: A General
Survey" • 2016: AlphaGo defeats the world champion in the Chinese
board game Go
• 1974: The first AI winter begins
• 2017: Sophia, the first robot to be granted citizenship
• 1980: Symbolics Lisp machines commercialized
• 2020: Curial GPT-3 LLM
• 1982: Fifth Generation Computer Systems
• 2022: ChatGPT
AI Applications Across Industries
• Healthcare: • Customer Service:
o Medical diagnosis and treatment planning o AI-powered chatbots and virtual
o Drug discovery and development assistants
o Natural language processing for
• Automotive (Self-Driving Cars): conversational interactions
o Perception, decision-making, and
navigation • Other Industries:
o Continuous learning and adaptation o Retail (recommendations, inventory
management)
• Finance: o Manufacturing (predictive maintenance,
o Fraud detection and risk management quality control)
o Stock trading analysis and portfolio o Entertainment (content
management recommendation, gaming AI)
UNDERSTANDING
DATA IN AI
Data - The Fuel for AI, Data Science, and ML
• Data is the foundational element that drives the development and performance
of AI systems. Without access to large amounts of high-quality data, AI
algorithms cannot learn and improve effectively.
• The quality, quantity, and diversity of data used to train AI models play a crucial
role in determining their accuracy, reliability, and generalizability.
o Quality: Data should be accurate, complete, and free from errors or inconsistencies.
o Quantity: AI models typically require large volumes of data to learn patterns and make
reliable predictions.
o Diversity: Data should cover a wide range of scenarios and edge cases to ensure the AI
system can perform well in various situations.
• AI algorithms, particularly those based on machine learning techniques, use data
to identify patterns, make predictions, and improve decision-making capabilities
over time.
Types of Data
• Structured Data: This type of data is organized in a predefined format, such
as databases or spreadsheets, making it easy to store, process, and analyze.
• Unstructured Data: This data lacks a predefined structure and can come in
various forms, such as text (emails, documents), images, audio, and video.
o Unstructured data often requires additional preprocessing and feature extraction
techniques to be used effectively in AI applications.
• Data Sources:
o Internet of Things (IoT) devices: Sensors and devices connected to the internet
generate vast amounts of data about their surroundings, usage patterns, and more.
o Online platforms: Social media, search engines, e-commerce websites, and other
online platforms are rich sources of user-generated data.
o Sensors and monitoring systems: Various sensors and monitoring systems collect
data from physical environments, machinery, and processes.
Ethical Data Use in AI
• As AI becomes more common and relies on data, it's vital to think about how we
handle information.
• Privacy concerns: When personal or sensitive data is used in AI, it can raise
worries about how that information is protected and if it's being used
appropriately.
• Being responsible: We need to manage data in a way that's fair and
accountable, which means keeping it secure, following privacy rules, and using it
ethically.
• Key principles: It's important to be transparent about how data is used, ensure
fairness in decision-making, and take responsibility for the outcomes of AI use.
INTRODUCTION TO
MACHINE LEARNING
Machine Learning(ML):
Empowering Systems • Machine learning is a subset of
artificial intelligence that enables

to Learn computers to learn from data and

experience without being explicitly
programmed.
• Machine Learning Algorithm can be
broadly classified into three types:
o Supervised Learning
o Unsupervised Learning
o Reinforcement Learning
• Machine learning powers
applications like facial recognition,
spam filtering, and self-driving cars.
Types of Machine Learning
• Supervised Learning is trained using the labeled dataset. Once the training
and processing are done, the model is tested by providing a sample test data to
check whether it predicts the correct output. Supervised learning can be divided
further into two categories of problem:
o Classification: In classification, the goal is to categorize data into predefined classes or categories based on its
features. It's like sorting data into different buckets or groups.
o Regression: Regression, on the other hand, deals with predicting continuous numerical values based on input data.
It's about understanding the relationship between variables to make predictions about outcomes.

• Unsupervised Learning is trained using the unlabeled dataset that is not

classified, nor categorized. In unsupervised learning, the model doesn't have a
predefined output, and it tries to find useful insights from the huge amount of
data
• Reinforcement learning is all about making decisions sequentially. In simple
words, we can say that the output depends on the state of the current input and
the next input depends on the output of the previous input
Applications of Machine Learning
• Recommendation Systems: An example of this is Netflix recommending
movies or TV shows based on your viewing history and preferences. Similarly,
Amazon suggests products based on your past purchases and browsing behavior.
• Image Recognition: Google Photos uses machine learning to recognize faces in
your photos and group them together. Also, self-driving cars use image
recognition to identify pedestrians, traffic signs, and other vehicles on the road.
• Natural Language Processing (NLP): Chatbots like Siri or Google Assistant
utilize NLP to understand and respond to natural language queries. Additionally,
spam email filters use NLP to identify and filter out unwanted emails.
• Predictive Analytics: An example is the use of machine learning in healthcare
to predict patient outcomes or diagnose diseases. For instance, machine learning
algorithms can analyze medical records to predict the likelihood of a patient
developing a particular condition.
ALGORITHMS
Statistical Foundations for Machine Learning
• Machine learning relies on statistical analysis to make sense of data.
• Key statistical concepts include:
o Regression Analysis: This technique finds the relationship between variables, often visualized as a
regression line.
o Correlation: This measures the strength and direction of the linear relationship between two
variables
o Mean:
▪ The average value of a set of numbers.
▪ Add up all the numbers and divide by the total count.
o Median:
▪ The middle value in a sorted list of numbers.
▪ If there's an even number of values, it's the average of the two middle numbers.
o Standard Deviation:
▪ A measure of how spread-out numbers are in a dataset.
▪ It shows the average distance of each data point from the mean.
o Standard Error:
▪ It measures the variability of a sample statistic.
▪ Shows how much the sample statistic (like the average) varies from the true population value.
o Mean Square Error (MSE):
▪ A measure of the average squared difference between predicted and actual values.
▪ It helps assess the quality of a model's predictions.
o T-Statistic:
▪ A measure of how much difference there is between groups, considering how much the data vary.
▪ It's used in hypothesis testing to determine if observed differences are significant.
o P-Value:
▪ The probability of observing a result as extreme as the one obtained, assuming the null hypothesis
is true.
▪ It helps decide whether observed results are statistically significant.
o Dependent Variable:
▪ The outcome being measured or observed in an experiment or study.
▪ Its value depends on other factors, including the independent variable.
o Independent Variable:
▪ The variable that is manipulated or controlled in an experiment.
▪ It's the variable hypothesized to have an effect on the dependent variable.
Correlation
• Correlation measures the strength and direction of the relationship between two
variables.
• Value ranges from -1 to 1,
• -1: Perfect negative correlation (as one goes up, the other goes perfectly down - think ice
cream sales vs. rainy days).
• 0: No correlation (changes in one don't affect the other - think shoe size vs. favorite color).
• 1: Perfect positive correlation (as one goes up, the other goes perfectly up - think exercise
vs. overall health).
• Correlation helps in understanding how changes in one variable are associated
with changes in another variable.
Linear Algorithm (Statistical) - Linear Regression
• Linear regression is a statistical method used for modeling the relationship
between a dependent variable and one or more independent variables.
• It assumes a linear relationship between the independent variables
(predictors) and the dependent variable (outcome).
• The goal is to find the best-fitting straight line that represents this
relationship, minimizing the difference between observed and predicted
values.
• Example: Predicting house prices based on size: The size of the house would
be the independent variable, and the price would be the dependent variable.
Linear regression helps find the equation of the line that best fits this
relationship, allowing prediction of house prices based on size.
Non-Linear Algorithm (Non-statistical) - Decision
Trees
• Decision trees make predictions by learning decision rules inferred from
the features of the data.
• They recursively partition the feature space into smaller regions based
on feature values, creating a tree-like structure of decision nodes and
leaf nodes.
• Each decision node represents a decision based on a feature value, while
each leaf node represents a predicted outcome.
• Example: Predicting product purchases based on age and income:
Decision nodes might include questions like "Is the person's age less
than 30?" and "Is the person's income greater than $50,000?" Leaf nodes
represent predicted outcomes, like "Will buy" or "Won't buy".
Conclusion
• History of AI and applications
• Understanding data, statistics, and heuristics is essential for success in AI, Data
Science, and ML.
• Importance of Data in AI.
• Machine Learning types and applications
• Regression analysis, decision trees, and correlation are important tools for
analyzing data and building predictive models.
• With the right knowledge and techniques, we can unlock the full potential of AI
to solve complex problems and drive innovation.
Q&A
THANK
YOU