AI

What is machine vision? Describe the components of machine

Machine vision is a field of artificial intelligence that enables computers to "see" and interpret images or videos from the real world. It involves using computer algorithms to extract meaningful information from visual data, such as objects, patterns, and movements.

Components of Machine Vision:

Image Acquisition: capturing images or videos.

Image Processing: enhance quality, such as removing noise or adjusting contrast.

Feature Extraction: such as edges, corners, or textures.

Object Recognition: This involves identifying and classifying objects within the images.

Motion Analysis: This involves tracking the movement of objects over time.

Decision Making: making decisions based on the information extracted from the images, such as controlling a robot or triggering an alarm.

What is artificial neural network? Then:
a. Discuss how back propagation algorithm is used to train ANN.
b. What is deep learning.

An Artificial Neural Network (ANN) is a computational model inspired by the way biological neural networks work. It consists of layers of interconnected nodes (neurons) that process information in parallel, mimicking the human brain.

a. Back Propagation Algorithm for Training ANN:

• Initialization: Assign random weights to connections.
• Forward Propagation: Calculate outputs using input data and activation functions.
• Error Calculation: Compute the error (difference between expected and actual outputs).
• Backward Propagation: Adjust weights to minimize the error by calculating gradients using the chain rule.
• Iteration: Repeat until the model converges to a desired accuracy level.

b. What is Deep Learning? Deep learning is a subset of machine learning that uses ANNs with many layers (deep networks). These layers extract hierarchical features automatically from raw data, making deep learning suitable for tasks like image recognition, speech processing, and natural language understanding.

What is expert system? Explain the major components of expert system.

a computer program that emulates the decision-making ability of a human expert in a specific domain. It uses a knowledge base of rules and facts to solve problems and make decisions.

Components of Expert System:

1. Knowledge Base: This component stores the domain-specific knowledge, including facts and rules.
2. Inference Engine: This component uses the knowledge base to reason about problems and make decisions.
3. User Interface: This component allows users to interact with the expert system, providing input and receiving output.
4. Explanation System: Provides reasoning for conclusions made by the system.

Differentiate Supervised Learning from Unsupervised Learning.

Supervised Learning: Involves training a model on labeled data, where each data point has a corresponding output or target value. The goal is to learn a mapping function that can predict the output for new, unseen data.  Common techniques: Classification, Regression.

Unsupervised Learning: Involves training a model on unlabeled data, where the goal is to discover underlying patterns, structures, or relationships within the data. Common techniques: clustering and dimensionality reduction.

    What do you mean by Artificial Intelligence? Then:
a. Define its goals.
b. What is Weak AI and Strong AI ?
c. Difference between Natural (Human) Intelligence & Artificial Intelligence.
d. Define the role of machine intelligence in the human life.

Artificial Intelligence (AI) is the simulation of human intelligence in machines programmed to think and act like humans.

a. Goals of AI:

1. Automate routine tasks.
2. Enhance decision-making.
3. Enable machines to learn and adapt.
4. Improve human-computer interaction.

b. Weak AI vs Strong AI:

• Weak AI: Systems designed for specific tasks (e.g., chatbots).
• Strong AI:  AI capable of general reasoning and problem-solving like humans.

c. Natural vs Artificial Intelligence:

Natural Intelligence:Biological brain, Adapts through experience, self-awareness, and emotional intelligence.

Artificial Intelligence: Machines/computers, Needs training data, 

d. Role of Machine Intelligence in Human Life:

Machine intelligence has the potential to improve many aspects of human life, such as:

• Healthcare: developing new treatments, and providing personalized care.
• Transportation: Enabling autonomous vehicles and improving traffic flow.
• Business: Automating tasks, improving customer service, and gaining insights from data.
• Science: Accelerating research in fields such as medicine, materials science, and climate change.

So, Machine intelligence helps in:

1. Automating complex tasks (e.g., surgery).
2. Enhancing productivity and efficiency.
3. Providing insights through data analysis.

What is an intelligent agent? Give short notes about multi-agent systems

Intelligent Agent: It is a smart entity that recognizes its environment, makes decisions, and acts to achieve specific goals. 

It can learn, adapt, and interact with other agents and the environment.

Examples: robots, software bots.

Multi-Agent Systems (MAS):

Consist of multiple interacting intelligent agents that can cooperate or compete with each other to achieve their goals. 

Used in distributed problem-solving (e.g., smart grids, swarm robotics).

Features: cooperation, coordination, and negotiation among agents.

Explain the difference between predictive and descriptive models of data mining?

Predictive Models:

    * Aim to forecast future events or outcomes based on historical data.

    * Used to make predictions, such as predicting stock prices.

Descriptive Models:

    * Aim to summarize, describe the characteristics of the data, and discover patterns in the data.

    * Used to gain insights into the data, such as identifying customer segments or understanding market trends.

Explain the main idea of KNN techniques; give an example of using it in classification?

K-Nearest Neighbors (KNN) is a simple algorithm used for classification. It classifies a data point based on the majority vote of its nearest neighbors.

Steps:

1. Calculate distances between the target point and other points.
2. Select the $k$-nearest neighbors.
3. Assign the majority class label to the target point.

Imagine you want to classify a new email as spam or not spam. You have a dataset of emails that have been labeled as spam or not spam. You can use KNN to classify the new email based on the similarity of their features (e.g., words, sender address, subject line).

Explain autonomy and adaptivity as key concepts for explaining AI?

• Autonomy: Ability to operate independently, make decisions, and take actions without human intervention. Example: self-driving cars.
• Adaptivity: Ability of an AI system to learn from experience and adjust its behavior to improve performance. Example: Recommendation systems, like in Netflix and Amazon.

Distinguish between realistic and unrealistic AI (science fiction vs. real life)?

Unrealistic AI (Science Fiction)

• Sentient AI: AI with self-awareness, and human-like emotions.   
• Superintelligence: AI that surpasses human intelligence in every aspect.   
• Mind Uploading: Transferring a human mind into a digital substrate.   
• AI Taking Over the World: AI seeks to dominate humanity.

Realistic AI (Real Life)

• Machine Learning: AI systems that learn from data and improve their performance over time.   
• Natural Language Processing (NLP): AI that can understand and generate human language.
• Computer Vision: AI that can interpret and understand visual information from the world.   
• Robotics: AI-powered robots that can perform tasks in the real world.

Express the basic philosophical problems related to AI including the implications of the Turing test and Chinese room thought experiment?

The Turing Test

Proposed by Alan Turing, the Turing Test is a method of determining whether a machine can have intelligent behavior equivalent to human. If a human cannot tell the difference between the responses of a human and a machine, then the machine is considered to be intelligent.

Chinese Room Test

Philosopher John Searle proposed the Chinese Room argument to challenge the idea that a machine can truly understand language. In this thought experiment, a person inside a room receives Chinese characters, follows a set of rules to manipulate them, and produces appropriate responses. While the person can successfully simulate understanding Chinese, they do not actually understand the language.  

Formulate a real-world problem as a search problem?

If we need to move from City A to City B, we have many ways and need to search for the best choice in terms of time, effort, cost, or other criteria. 

Many problems can be framed as search problems, where the goal is to find a sequence of actions that leads from an initial state to a desired goal state.

Key points:

• Problem Formulation: Defining the problem, including the initial state, goal state, and the possible actions that can be taken.
• Search Techniques: Using algorithms to explore the possible solutions and find the optimal path.
• Static Environments: Problems where the environment doesn't change while the agent is making decisions.

Formulate a simple game (such as tic-tac-toe) as a game tree?

To find the next move, AI needs to explore all possible moves and their consequences, helping to make optimal decisions. concept of a game tree helps in representing and analyzing games in AI.

Which includes:

• Nodes and Levels: Nodes represent game states, and levels correspond to each player's turn.
• Root Node: The initial state of the game.
• Child Nodes: Possible states resulting from a player's move.
• Game Termination: The tree ends in terminal nodes, representing game-over states (win, loss, or draw).