Introduction to Artificial Intelligence and Data Science
ID-142
Anup Kumar Halder, PhD,
Dept. CSE, IIIT Raichur
25th November 2024
Intelligence
Intelligence can be defined as the ability to learn and perform suitable techniques to
solve problems and achieve goals, appropriate to the context of an uncertain, ever-varying
world.
A fully pre-programmed factory robot is flexible, accurate, and consistent but not
intelligent.
What is Artificial Intelligence (AI)?
Artificial Intelligence (AI) is a branch of computer science focused on creating systems
that can perform tasks requiring human intelligence, such as reasoning, learning, and
decision-making.
Historical Perspective
• Coined by John McCarthy in 1955.
• Defined as “The science and engineering of making intelligent machines.”
Core Characteristics of AI
1. Learning: Acquiring knowledge from data or experience.
2. Reasoning: Making decisions based on logical rules.
3. Problem-Solving: Identifying and executing steps to achieve goals.
4. Perception: Interpreting sensory inputs like images or sounds.
1
Machine Learning (ML): The Heart of AI
Machine Learning (ML) is a subset of AI focused on algorithms that allow machines to
learn from data without explicit programming.
Machine Learning is the part of AI studying how computer agents can improve their
perception, knowledge, thinking, or actions based on experience or data.
Types of Machine Learning
1. Supervised Learning:
• Learns from labeled datasets.
• Example: Predicting house prices based on features like size, location.
2. Unsupervised Learning:
• Finds patterns in unlabeled data.
• Example: Clustering customers into groups based on shopping behavior.
3. Reinforcement Learning:
• Agents learn optimal actions through trial-and-error to maximize rewards.
• Example: Learning to play chess or controlling robots.
Tools and Technologies
Data Science Tools
• Programming Languages: Python, R, SQL.
• Visualization: Matplotlib, Seaborn, Tableau.
• Big Data Frameworks: Apache Hadoop, Spark.
AI Tools
• Machine Learning Libraries: Scikit-learn, TensorFlow, PyTorch.
• Natural Language Processing: SpaCy, NLTK, Hugging Face.
• Deep Learning Frameworks: TensorFlow, PyTorch.
2
Figure 1: Venn diagram representing the hierarchical and overlapping relationships be-
tween Artificial Intelligence (AI), Machine Learning (ML), Deep Learning (DL), and Data
Science (DS): Artificial Intelligence (AI): The overarching field that involves creating
systems capable of intelligent decision-making and problem-solving.
Machine Learning (ML): A subset of AI focusing on algorithms that allow systems to
learn from data without explicit programming.
Deep Learning (DL): A specialized subset of ML using neural networks to solve com-
plex tasks like image recognition and natural language processing.
Data Science(DS): An interdisciplinary field that leverages AI, ML, and statistical
techniques to extract insights and knowledge from data.
Data Science
Data science is an interdisciplinary field that uses scientific techniques from statistics
and computer science to systematically extract knowledge from data.
Data science is the collection, analysis, and interpretation of data to drive in-
sight and solutions to a given problem. It involves skills from several disciplines, in-
cluding mathematics, computer science, algorithms, statistical analysis, and artificial
intelligence/machine learning.
3
Relationship Between Data Science and AI
• Overlap:
– AI relies on large volumes of data for training.
– Data Science uses AI techniques (e.g., machine learning) for predictions.
• Differences:
– AI focuses on automation and decision-making.
– Data Science emphasizes extracting insights and understanding data.
Applications of Data Science and AI
• Healthcare:
– Diagnosing diseases using medical imaging (AI).
– Predicting outbreaks using historical data (Data Science).
• Finance:
– AI: Fraud detection using anomaly detection.
– Data Science: Risk assessment for loans.
• Retail:
– AI: Personalized recommendations using deep learning.
– Data Science: Analyzing sales trends for better inventory management.
• Transportation:
– AI: Autonomous vehicles and traffic prediction.
– Data Science: Optimizing delivery routes.
Ethical Considerations in Data Science and AI
• Bias: Algorithms trained on biased data can make unfair decisions.
• Privacy: Ensuring data is used responsibly and securely.
• Accountability: Clear responsibility for decisions made by AI systems.
4
Future Trends in Data Science and AI
• Explainable AI: Making AI models transparent and understandable.
• Edge AI: Deploying AI models on devices like smartphones for real-time process-
ing.
• Multimodal AI: Combining different data types (e.g., text, images) for better
insights.
• Automated Machine Learning (AutoML): Simplifying the process of building
models.
Questions:
• What type of learning system is adopted by ChatGPT? Explain in detail.
