✅ AI Reflection, Project Cycle and Ethics – Final Notes (Class 9)
📘 1. Definition of AI (Artificial Intelligence)
AI stands for Artificial Intelligence.
It helps machines to:
o Solve problems
o Adapt to new situations
o Make data-driven decisions
AI is capable of:
o Speech recognition
o Solving problems
o Making predictions
o Taking decisions without human help
📱 2. Applications of AI in Smartphones (Computer Vision)
These smartphone features use Computer Vision:
1. Face unlock
2. Google Lens
3. Augmented Reality filters in social media apps (e.g., Snapchat,
Instagram)
4. Barcode and QR code scanners
5. Image search in photo gallery
🔁 3. Three Domains of AI and Their Data Types
(i) Natural Language Processing (NLP)
Works with textual and spoken data
Helps machines to understand, generate, and manipulate human
language.
(ii) Computer Vision (CV)
Works with visual data – images and videos.
Helps in understanding and interpreting visual content.
(iii) Data Science
Works with numerical, statistical, and structural data
Helps find patterns and insights from large datasets.
Diagram: Domains of AI
┌───────────────┐
│ Domains of AI│
└─────┬─────────┘
│
┌─────────┴──────────┐
│ │
CV NLP
│ │
└─────────┬──────────┘
↓
Data Science
📌 4. Stages of AI Project Cycle
1. Problem Scoping
o Identify and define the problem clearly.
o 🔹 Example: “How to improve crop yield using AI”
2. Data Acquisition
o Collect necessary and relevant data.
o 🔹 Example: Weather, pest attack, soil reports, etc.
3. Data Exploration
o Analyze data and find patterns.
o 🔹 Example: How weather affects crops.
4. Modelling
o Build a machine learning model.
o 🔹 Example: Predict the best time to use pesticide.
5. Evaluation
o Test the model for accuracy.
o 🔹 Example: How well does it predict crop yield?
6. Deployment
o Use the trained AI model in a real-life setting.
o 🔹 Example: Deploying the AI tool to help farmers in real time.
📊 5. Types of Data
Training Data – Raw data used to train the AI model.
Testing Data – Processed data used to test if the model works well.
📈 6. Importance of Data Visualization
Helps us understand complex data using charts and graphs
Makes patterns and trends clear
Helps in finding errors or outliers
Supports smart decisions with clarity
⚙️7. AI Modelling
AI modelling is the process of teaching a machine to make decisions or
predictions using data and examples.
AI model:
o Learns from data
o Solves real-life problems
(e.g., identify faces or predict weather)
We can use:
o Own model
o Pre-existing model
📚 8. Homework Questions
Q1. If your AI model is giving incorrect predictions, how will you check
whether the problem is in the model or in the data?
➤ Steps to check:
See if data is incomplete or biased
Check if model is trained on correct examples
Q2. What are some consequences of using incomplete or biased data to
evaluate an AI model?
➤ Consequences:
AI may give wrong or unfair results
AI may become biased
Creates ethical issues
🤖 9. AI, ML and DL – Explained
Term Description
AI (Artificial
Machine's ability to think and act like humans
Intelligence)
Term Description
ML (Machine A part of AI that lets machines learn from data without
Learning) programming
A special type of ML using neural networks to solve
DL (Deep Learning)
complex tasks
🧠 10. Evaluation (Extra Notes)
Final stage of the AI project cycle.
Check model performance and accuracy.
Purpose: Ensure AI meets the goal defined during Problem Scoping.
11. Deployment (Extra Notes)
Deploy means to put the trained AI model into a real-world app
The model becomes usable by people (e.g. a weather alert tool for
farmers)
