AI , ML and Data Science

Artificial Intelligence (AI)

Imagine you have a brain. You can:

• See things
• Hear and understand language
• Think and learn from experience
• Decide what to do

Now imagine we try to give these powers to a machine.   That is called Artificial Intelligence (AI).

• “Artificial” means man-made.
• “Intelligence” means ability to think, learn, and solve problems.
• AI (Artificial Intelligence) is the ability of a computer or machine to think, learn, and make decisions like humans.
• It uses data and algorithms to solve problems, recognize patterns, and perform tasks.
• AI is used in daily life, like voice assistants, self-driving cars, and recommendation systems.

Machine Learning (ML)

Machine Learning (ML) is a branch of Artificial Intelligence (AI) that allows computers to learn from data and improve their performance without being explicitly programmed.

Machine Learning = Teaching computers to learn from data and experience.

Focus on development of computer program that can access data use it learn themselves.

DEEP LEARNING

Deep learning is a type of artificial intelligence that uses layers of digital “neurons” to learn patterns from huge amounts of data. By processing lots of examples—like pictures or sounds—these systems teach themselves how to solve problems, often outperforming humans in tasks like image recognition or language translation. Instead of following step-by-step instructions, deep learning models gradually improve by finding features and rules on their own. It’s similar to how our brains learn from experience but happens much faster and with more data

Neural Network

A neural network is a computer system designed to work like the human brain. It has layers of connected units called neurons that learn from data by adjusting their connections. This helps the network recognize patterns and make decisions, such as identifying images or understanding speech. In short, it’s a smart system that learns and thinks like a brain to solve problems.

Types of Neural Network

1. Feed Forward Neural Network (FNN) – Data flows only forward (input → hidden → output).
2. Convolutional Neural Network (CNN) – Best for images, video, pattern recognition.
3. Recurrent Neural Network (RNN) – Works with sequential data (speech, text, time series).
4. Radial Basis Function Network (RBFN) – Uses distance-based activation, good for classification.
5. Modular Neural Network (MNN) – Several smaller networks work together.
6. Generative Adversarial Network (GAN) – Generates new data (like fake images, deepfakes).
7. Self-Organizing Map (SOM) – Used for clustering and visualization.
8. Autoencoders – For data compression and noise removal.
9. Long Short-Term Memory (LSTM) – A special type of RNN, remembers long-term patterns.

How AI ,ML and DATA SCIENCE WORK IN CORE BRANCHS ?

1. Mechanical Engineering

• AI/ML Uses:

1.     Predictive maintenance of machines using IoT sensors (detect failures before breakdown).

2.     Quality control with computer vision (detect defects in manufactured parts).

3.     ML models for optimization of design (lighter yet stronger materials).

• Data Science Uses:

1.     Analyzing huge datasets from machines (temperature, vibration, pressure).

2.     Simulation data analysis (CFD, FEA).

2. Civil Engineering

• AI/ML Uses:

1.     Smart traffic management using AI (predict congestion, optimize signals).

2.     ML for structural health monitoring (detect cracks, stress points in bridges/buildings).

3.     Predicting construction project delays & cost overruns.

• Data Science Uses:

1.     Analyzing geotechnical survey data for foundation strength.

2.     Studying climate and soil data for construction planning.

3. Electrical Engineering

• AI/ML Uses:

1.     Smart grids: predicting electricity demand and supply.

2.     Fault detection in power systems using ML.

3.     AI-based load balancing for efficient energy usage.

• Data Science Uses:

1.     Processing electrical consumption data to optimize tariff plans.

2.     Analyzing renewable energy patterns (solar/wind).

4. Electronics & Communication

• AI/ML Uses:

1.     Signal processing with AI (noise reduction, pattern recognition).

2.     ML in 5G/6G networks for optimal bandwidth allocation.

3.     AI-powered chip design & testing.

• Data Science Uses:

1.     Analyzing communication logs for performance.

2.     IoT device data monitoring.

5. Computer Science & IT (already direct field)

• AI/ML Uses:

1.     Natural Language Processing (chatbots, translation).

2.     Computer Vision (face recognition, self-driving cars).

3.     Deep Learning for cybersecurity.

• Data Science Uses:

1.     Big Data analytics for business and research.

2.     Cloud-based AI solutions.

 6. Chemical Engineering

• AI/ML Uses:

1.     Process optimization (temperature, pressure control in reactors).

2.     Predictive modeling of chemical reactions.

3.     Safety monitoring (detecting gas leaks via AI sensors).

• Data Science Uses:

1.     Large-scale process data analysis.

2.     Studying material properties with molecular data.

Data Science

Data science is the study of data to find useful insights and patterns that help solve real-world problems or guide business decisions. It combines math, statistics, and computer programming to analyze and make sense of large amounts of data. By discovering hidden trends, data science helps companies predict outcomes and improve services. In short, it's all about turning raw data into smarter decisions and actions.

Types of Data

• Nominal Data: Labels or names without any order, like color or gender.
• Ordinal Data: Labels with a specific order, like ratings or education levels.
• Discrete Data: Countable numbers, like number of students or cars.
• Continuous Data: Measurable values that can take any value within a range, like height or temperature.

·        Categorical Data: Categorical data refers to data that can be divided into groups or categories. These categories represent qualities or characteristics, not numerical values.

Types of Machine Learning

1. Supervised Learning

• Learns from labeled data (input + correct output given).
• Goal: Prediction.
• Example: Predicting house prices, spam email detection.

2. Unsupervised Learning

·        Learns from unlabeled data (only input, no correct output).

·        Goal: Find patterns, groups.

·        Example: Customer segmentation, market basket analysis.

3. Semi-Supervised Learning

• Mix of labeled + unlabeled data.
• Example: Medical image classification (few labeled, many unlabeled).

4. Reinforcement Learning

• Learns by trial and error with rewards & penalties.
• Example: Self-driving cars, game-playing AI (Chess, Go).

Fig . Difference between supervised and Unsupervised learning

Regression Analysis

• A statistical & ML method used to find the relationship between variables.
• It predicts a dependent variable (Y) based on one or more independent variables (X).

 Types of Regression

1.     Linear Regression - Straight-line relation (e.g., study hours → marks).

2.     Multiple Linear Regression - Many factors affect output (e.g., price depends on size, location, age of house).

3.     Polynomial Regression - Curved relation.

4.     Logistic Regression - For classification (yes/no, pass/fail).