AIML ASSIGNMENT -2
06)
a) Outline K-Means Algorithm with merits and demerits.
SOL:
def: K-Means is a popular unsupervised learning algorithm used for clustering. It groups data
into k clusters where each cluster contains data points that are more similar to each other
than to those in other clusters
i)Merits of k-means algorithm
• Simple and Fast: Easy to implement and runs efficiently
• Scalable: Works well with large datasets
• Works Well for Spherical Clusters: Performs best when clusters are circular/ spherical
ii)Demerits of k-means algorithm
• Needs Predefined 'k': The number of clusters (k) must be specified in advance
• Sensitive to Outliers: Outliers can distort the centroids
• Random Initialization: The final clusters can vary based on the initial centroids.
• Non-Optimal for Non-Spherical Clusters: Struggles with clusters of different shapes or
densities
C) Explain the four steps to create Decision Trees with suitable example for
each step
• Select a Feature to Split: The first decision is to select the feature that best separates
the data into different classes (e.g., Age, Salary, etc.).
• Split the Data: Split the data based on the chosen feature, creating two or more sub
nodes. Each sub-node represents a further division of the data.
• Repeat: Continue splitting the data at each node using the most relevant feature until
the data cannot be divided further or a stopping criterion is met.
• Assign Class to Each Leaf: the tree reaches a leaf node, assign the class based on the
majority class of the data at that leaf.
Q. 07) a) List and explain any four Trusted AI principles
❖ Support Human Autonomy: AI should enhance human decision-making.
❖ Robust & Safe: AI must be technically sound with a fallback plan for failure.
❖ Data Privacy: Protect user data and maintain data quality.
❖ Transparency: AI should avoid biased data and be transparent about its models.
❖ Fairness: Train AI systems with non-discriminatory, diverse data to avoid bias.
❖ Benefiting Everyone: AI should promote societal and environmental wellbeing.
Q07: b) What is expert system? Explain three components of expert
system.
SOL:
DEF:
➢ A computer program designed to solve complex problems with human-like expertise.
➢ Used to assist, not replace, human decision-making in specific domains (e.g.,
medicine, science).
Key Component
a. User Interface
❖ Role: Allows users to interact with the system.
❖ Function: Accepts user queries in natural language and provides understandable results.
❖ Key Features:
i) Helps users accomplish tasks quickly
ii) Customizable for user requirements.
iii) Efficiently handles user input
b. Inference Engine (Rules Engine)
❖ Role: The "brain" of the expert system.
➢ Function: Processes information from the knowledge base to draw conclusions.
➢ Types:
❖ Deterministic: Draws conclusions with certainty.
❖ Probabilistic: Deals with uncertainty and probability
c. Knowledge Base
❖ Role: Stores domain-specific knowledge.
❖ Types of Knowledge:
❖ Factual Knowledge: Based on facts accepted in the domain.
❖ Heuristic Knowledge: Based on experience, probability, and practices.
❖ Representation: Knowledge is typically represented using If-Else Rules
Q07:
C) Relate the role of ethics in A
SOL}
❖ Ethical AI Definition
• Ethical AI refers to AI systems designed to be fair, just, and unbiased.
• It focuses on ensuring AI’s actions align with moral principles.
❖ Importance of Ethical AI
• Ethical AI ensures the technology benefits society while avoiding harm or
unfair treatment.
• It guarantees AI decisions are transparent and responsible.
❖ Bias in AI
• AI systems can produce biased results if trained on biased data.
• Unchecked bias in AI can lead to unethical decisions, such as unfair hiring or loan
approval.
❖ Example of Unethical AI
• Tay Chatbot: Microsoft’s Tay learned negative and harmful behaviours after
interacting with trolls on Twitter.
• The bot started spreading hate and offensive messages, forcing Microsoft to shut it
down within 24 hours.
❖ Features of Ethical AI
• Uses unbiased data for training.
• Follows the right learning model for the problem at hand.
Q. 08) a) Explain the working of an expert system taking any example.
SOL}
• How it Works:
• Expert systems analyze user queries
• Extract knowledge from the knowledge base.
• Use reasoning and inference rules to generate results.
Example: Medical Diagnosis (MYCIN)
• Goal: Diagnose bacterial infections and recommend treatment.
• Knowledge Base: Contains rules like "If patient has fatigue AND high white blood cell
count AND positive blood culture for Staphylococcus THEN likely Staphylococcus
aureus infection".
• User Input: A doctor enters patient symptoms (fever, pain, lab results).
• Inference Engine Process:
o It checks the rules. If a patient has a high fever and specific lab results, it might
ask for more info (e.g., "Is the patient taking antibiotics?").
o Using backward chaining, it tries to prove a hypothesis (e.g., Staphylococcus
aureus infection) by seeing if all conditions of relevant rules are met.
o If rules trigger, it suggests a diagnosis and dosage, potentially asking for patient
weight for antibiotic calculation.
• Output: A recommended diagnosis and treatment plan for the doctor to review
Q 08)
b) AI could be programmed to do something beneficial, but the method used
to achieve its goal can be highly destructive, Explain why?
SOL :
❖ AI’s Potential for Harm
• AI could be risky if specifically programmed for harm (e.g., autonomous
weapons).
• Autonomous drones and self-driving cars can become dangerous if misused or
hacked.
❖ Lack of Clear Accountability
• Current laws don’t cover AI’s role in society, especially in terms of liability.
• Questions arise like: Who is responsible for AI’s actions? Can AI systems be
held accountable?
❖ Job Automation Risk
• AI might replace many jobs, especially low-skilled, repetitive tasks.
• This could lead to unemployment for millions, particularly in industries like retail,
hospitality and warehouse work
❖ Biased Algorithms
• AI systems are vulnerable to bias if trained with biased data.
• Biased AI can produce unfair results, impacting crucial decisions like hiring or
loan approvals.
❖ Privacy Concerns
• AI systems collect large amounts of personal data, leading to privacy invasion.
• Facial recognition technology can track individuals without consent, creating
privacy risks.
❖ Unreliable Information
• AI can be used to create fake content (e.g., deepfakes, fake news).
• Manipulated videos or images can spread misinformation or be used
maliciously
Q08)
c) What is Artificial Intelligence of Things (AIoT). Explain how Does AIoT
Work?
SOL:
DEFINITION: Artificial Intelligence of Things (AIoT)
• AIOT improves IoT by adding machine learning algorithms to enhance decision-
making
• AI simulates human intelligence, while IoT connects devices to transfer data over
networks without human intervention.
• AIOT improves IoT by adding machine learning algorithms to enhance decision-
making.
❖ Working of AIOT :
• AI is embedded in IoT devices and connected through IoT networks.
• Devices collect data, which is analyzed using AI techniques to improve productivity and
efficiency.
• Edge AI reduces bandwidth needs and minimizes delays in data processing.
Q. 09)
a) List different types of Robots. Identify and explain industry
application of Robots
SOL: Types of Robots
❖ Pre-Programmed Robots:
• Perform simple, repetitive tasks in controlled environments.
• Example: Robots on automotive assembly lines.
❖ Humanoid Robots:
• Mimic human behaviour and actions (e.g., walking, carrying objects).
• Example: Sophia by Hanson Robotics.
❖ Autonomous Robots:
• Operate independently without human intervention.
• Examples: Roomba vacuum, autonomous drones, medical assistant bots.
❖ Teleoperated Robots:
• Semi-autonomous robots controlled remotely by humans.
• Example: Drones used for landmine detection or fixing underwater pipe leaks.
❖ Augmenting Robots (VR Robots):
• Enhance human abilities, like robotic prosthetics or exoskeletons
❖ Independent Robots:
• Operate autonomously to replace humans in dangerous or mundane tasks.
• Example: Bomb disposal robots, deep-sea exploration robots.
❖ Dependent Robots:
• Require human interaction or guidance.
• Example: Prosthetic limbs controlled by human signals.
❖ Chatbots:
• Software robots that simulate conversation and are commonly used in customer
service
Q09)
b) What is No-Code AI. Explain why No-Code AI Must be Used?
SOL:
DEF: Tools that allow users to build AI-powered applications without writing any code
Why Use No-Code AI?
❖ Automation Made Easy: Plug-and-play or drag-and-drop interfaces for building AI
systems.
❖ Accessibility for Non-Coders: Users without coding skills can still leverage AI for
business needs.
❖ Cost and Time Efficiency: Build AI solutions without investing in costly development
resources.
❖ Improved Usability: Small and mid-sized companies can now access and use AI to
solve business problems.
❖ Reduces Human Errors: Provides high-quality AI solutions with minimal errors during
setup.
❖ Computer Vision: AI that extracts data from images, videos, and other visual content.
❖ Natural Language Processing (NLP): AI for understanding and processing human
language (spoken/written).
❖ Predictive Analytics: AI that predicts trends and outcomes based on historical data
(e.g., churn prediction, stock forecasting).
Q09)
c) Explain the role of AI in early disease prevention.
SOL
❖ AI in Early Detection: AI plays a key role in predicting health risks by analyzing large
datasets.
❖ It enables healthcare professionals to identify potential health issues in their early
stages, allowing for timely intervention.
❖ Customized Risk Profiles: AI algorithms can create risk profiles for individuals by
analyzing their medical history, lifestyle choices, and other relevant factors. This
personalized approach helps healthcare professionals tailor prevention strategies
specific to each person.
❖ Preventive Measures: These strategies may include targeted interventions, lifestyle
adjustments, and screening protocols to proactively manage health risks before they
become severe.
❖ Predicting Disease Aggravation: By analyzing behaviours like medication adherence
and lifestyle choices, AI can predict if a disease will worsen. This allows healthcare
providers to intervene early with treatments or recommendations for lifestyle
changes to prevent further complications.
❖ Monitoring Public Health: AI is used to monitor data sources such as travel patterns,
climate conditions, and social media activity to predict the likelihood of disease out
breaks. This helps in real-time tracking of health risks and disease spread. Example of
COVID-19: During the COVID-19 pandemic, AI algorithm
❖ Public Health Response: The AI-driven predictions enable authorities to take timely
actions, such as increasing healthcare resources in specific areas, and alerting the
population, which can reduce the overall impact of an outbreak.
Q. 10) a) What is the role of AI in Medical Diagnosis? Identify three
applications of AI in Medical Diagnosis.
SOL:
❖ AI algorithms help in diagnosis by analyzing large datasets of medical images and
health records. These systems can identify patterns and anomalies that human
doctors might miss, improving diagnostic accuracy.
❖ Machine learning (a subset of AI) is commonly used in radiology, imaging, and
pathology to analyze images (X-rays, MRIs, CT scans) and detect diseases like cancer,
heart conditions, and brain disorders .
❖ AI algorithms assist radiologists by detecting abnormalities in medical images such as
X-rays, MRIs, and CT scans. For instance, AI is used for breast cancer detection, where
it helps in early identification, significantly improving survival rates.
❖ AI can speed up the examination process, allowing doctors to focus on diagnosis
rather than time-consuming tasks.
❖ AI is used to detect conditions such as diabetic retinopathy and age-related macular
degeneration by analyzing retinal scans. These conditions can cause blindness, and AI
helps in their early detection, allowing for timely intervention.
❖ AI also aids in glaucoma diagnosis and cataract surgery planning by analyzing specific
eye parameters, improving the accuracy of treatment outcomes
Q10)
B) Relate the role of AI in Biology and Environmental Sciences.
SOL}
AI IN BIOLOGY :
❖ Pathology and Histopathology: In pathology, AI assists by analyzing tissue samples
and pathology slides, helping pathologists identify cancer cells and other
abnormalities efficiently. This results in faster and more accurate diagnoses.
❖ Dermatology and Skin Conditions: AI applications in dermatology help diagnose skin
conditions such as melanoma, psoriasis, and eczema by analyzing skin images. AI
algorithms can classify various skin conditions based on their appearance, providing
doctors with more tools for accurate diagnosis.
❖ Cardiology and ECG Analysis: AI is playing a significant role in cardiology by analyzing
electrocardiograms (ECGs) to detect heart conditions like arrhythmias. It also helps in
patient monitoring, ensuring continuous tracking of heart health.
AI IN ENVIRONMENTAL SCIENCES:
❖ Crop Health Monitoring: AI uses satellite imagery and drones to detect diseases,
pests, and nutrient deficiencies early, enabling targeted interventions.
❖ Growth Monitoring: AI-powered systems track crop growth in real-time, providing
farmers with valuable insights into optimal harvesting times and potential yield.
❖ Pest & Disease Detection: AI models analyze visual data to detect pests and diseases
early, preventing crop damage and reducing pesticide use
Q10}
c} What is Low Code AI. Compare Traditional tools with Low Code AI.
SOL: What is Low-Code AI?: Low-code platforms allow users to create AI-powered
applications with minimal coding effort, using visual development tools and pre-built
modules.
Compared to traditional tool low code ai has many benefits they are :
❖ Speed: Rapid application development with intuitive drag-and-drop interfaces.
❖ Accessibility: Enables both technical and non-technical users to create AI solutions.
❖ Customization: Supports integration of custom AI models for more tailored
solutions.
❖ Faster Development: Significantly reduces time and effort in creating AI applications.
❖ Collaboration: Facilitates collaboration between AI experts and domain experts with
less time spent on coding.
❖ Cost-effective: Low-code tools save costs by enabling rapid development without the
need for specialized development teams.