Artificial Intelligence

CBSE Class 10 AI – Unit 4: Statistical Data | No-Code AI & Orange Data Mining

Class 10 · Artificial Intelligence

```

Unit 4: Statistical Data

Complete Unit Syllabus

  • 4.1 No-Code AI for Statistical Data
    • Meaning of No-Code AI
    • No-Code and Low-Code
    • Some No-Code Tools
  • 4.2 Statistical Data: Use Case Walkthrough
    • Important Concepts in Statistics
    • Orange Data Mining
    • AI Project Cycle in Orange Data Mining (Palmer Penguins Case Study)

Introduction to Statistical Data

Statistical Data refers to information that is collected, organised and analysed to identify patterns, relationships and trends.

In Artificial Intelligence, data plays a very important role. AI systems learn from data and use the patterns found in that data to make predictions or decisions.

Think of it like a school examination:

A school collects marks, attendance, participation and other academic information. By analysing this data, teachers can identify learning gaps, compare performance and make informed decisions.

Similarly, AI uses statistical data to discover patterns and make predictions.

This unit also introduces No-Code AI, which allows users to create AI and Machine Learning solutions using visual tools instead of writing complex computer programs.

4.1 No-Code AI for Statistical Data

What is No-Code AI?

No-Code AI is a way of building Artificial Intelligence models without requiring the user to write computer programming code.

Instead of writing lines of code in programming languages such as Python or Java, users can use visual interfaces, menus and drag-and-drop features to create AI projects.

Aspect No-Code AI
Coding Required No programming code is required.
Interface Uses visual interfaces and drag-and-drop components.
Users Can be used by people without programming expertise.
Purpose To make AI development easier and more accessible.

Goal of No-Code AI

The main goal of No-Code AI is to make Artificial Intelligence accessible to everyone, not only computer scientists and professional programmers.

People such as doctors, musicians, teachers, architects and business professionals can use No-Code AI tools to develop solutions for their specific requirements.

Key Idea:

No-Code AI reduces the technical barrier to Artificial Intelligence development.

Key Benefits of No-Code AI

  • No Programming Required: Users can create AI solutions without learning a programming language first.
  • Easy to Use: Visual interfaces make the process easier for beginners.
  • Faster Development: AI workflows can be created using drag-and-drop components.
  • Accessibility: Non-technical professionals can participate in AI projects.
  • Fewer Syntax Errors: Since users do not manually type program code, common programming syntax errors are avoided.

Real-Life Example – School Canteen App

Student Example

Imagine a student wants to create an application that predicts which snacks are likely to sell out during the school recess.

The student has historical canteen sales data but does not know Python or another programming language.

Using a No-Code AI tool, the student can create a prediction workflow using visual components instead of writing complex programming code.

The model can learn from previous sales data and help predict future demand.

Real-Life Example – Study Mood Analyzer

Teacher Example

A teacher wants to understand whether students appear relaxed or stressed before examinations.

A tool such as Teachable Machine can be used to train a simple AI model to recognise different categories of images, sounds or poses.

Such technology can help demonstrate how AI models can be developed without traditional programming.

Real-Life Example – Library Sorter

School Library Example

A school librarian wants to automatically identify different categories of books from their covers.

A No-Code image classification tool can be trained using examples of book covers and their corresponding categories.

The trained model can then classify new book-cover images based on the patterns it has learned.

No-Code, Low-Code and High-Code

AI and application development can broadly be approached using High-Code, Low-Code and No-Code methods.

Feature High-Code
(Custom Coding)
Low-Code No-Code
Meaning Traditional programming where developers write code manually. Uses visual tools but may require some manual coding. Uses visual tools and does not require coding.
Coding Extensive coding is required. Some coding may be required. No coding is required.
Cost Generally expensive because expert developers may be required. Generally less expensive than High-Code. Generally the least expensive option.
Customization Unlimited: developers can create highly customised solutions. Limited: customization depends on the platform. Fixed: users are generally limited to the functions provided by the tool.
Speed Slowest because development is largely manual. Faster than High-Code. Fastest because solutions can often be created using ready-made components.
Technical Skill High programming expertise is generally required. Some technical knowledge may be required. Little or no programming knowledge is required.

Easy Way to Remember

High-Code → Write the code yourself.

Low-Code → Mostly visual + some coding.

No-Code → Visual tools without coding.

Limitations of No-Code AI

Although No-Code AI is easy and fast, it also has some limitations.

  • Limited Flexibility: Users are restricted to the functions provided by the platform.
  • Limited Customization: Highly specialised features may not be possible.
  • Platform Dependency: Users depend on the capabilities and features of the selected platform.
  • Advanced Requirements: Complex AI applications may still require traditional programming.

Some No-Code AI Tools

Tool Description Important Feature
Orange Data Mining An open-source tool used for data visualization and Machine Learning. Visual canvas with connected widgets.
Teachable Machine A web-based tool that allows users to train AI models using examples. Can work with images, sounds and poses.
Lobe AI A simple tool for training custom image classification models. Easy visual model-training interface.
Azure Machine Learning A cloud-based Machine Learning service. Supports data preparation, model training and evaluation through visual tools.
Google Cloud AutoML A service designed to help users build Machine Learning models with limited AI expertise. Helps create customised Machine Learning models.

Tool Identification – Quick Revision

If the question mentions... Think of...
Open-source + visual data mining Orange Data Mining
Images, sounds or poses Teachable Machine
Custom image classification Lobe AI
Microsoft cloud Machine Learning Azure Machine Learning
Google Cloud customised Machine Learning Google Cloud AutoML

Case Study: Kayla the Zoo Dietitian

Case Study

Kayla manages food requirements for animals in a zoo.

The prices of meat and vegetables are increasing rapidly. The zoo wants to predict future price increases so that it can request appropriate funding from the government.

Kayla has expert knowledge of animal nutrition, but she does not know how to code.

Question 1

Which approach should Kayla choose to build her price prediction model? Justify your answer.

Answer:

Kayla should choose the No-Code AI approach, using a suitable tool such as Orange Data Mining.

Since Kayla does not have programming expertise, a No-Code tool allows her to create an AI workflow using visual and drag-and-drop components instead of writing complex code.

Question 2

Identify one disadvantage Kayla might face while using a No-Code approach.

Answer:

Lack of Flexibility.

Kayla may be limited to the features and functions provided by the selected No-Code platform. A highly specialised feature may not be possible without traditional programming.

Competency-Based Questions – 4.1

1. Multiple Choice Question

Q. Which terminology best describes a platform that allows a non-technical musician to build an AI model without writing scripts?

a) High-Code AI
b) Low-Code AI
c) No-Code AI
d) Manual AI

Answer: c) No-Code AI

Explanation: No-Code AI allows users to create AI solutions without writing programming code.

2. Assertion and Reasoning

Assertion (A): No-Code AI tools make Artificial Intelligence more accessible to the general public.

Reason (R): These tools rely on complex manual coding which ensures higher accuracy for non-technical users.

a) Both A and R are true and R is the correct explanation.
b) Both A and R are true but R is not the correct explanation.
c) A is true but R is false.
d) Both A and R are false.

Answer: c) A is true but R is false.

Explanation: No-Code tools use visual interfaces and drag-and-drop components rather than complex manual coding.

3. Tool Identification

Scenario: A student wants to create a Machine Learning project using an open-source platform with visual tools for data mining.

Q. Which tool would be most suitable?

Answer: Orange Data Mining.

Reason: Orange is an open-source data mining and Machine Learning tool that uses a visual workflow.

4. Application-Based Question

A school wants to develop a simple AI system but its teacher has no programming experience.

Q. Should the teacher immediately learn a programming language before starting the project?

Answer: Not necessarily. The teacher can first explore a suitable No-Code AI platform.

No-Code tools can provide visual components that allow beginners to build AI workflows without traditional coding.

4.2 Statistical Data: Use Case Walkthrough

Statistical concepts help us understand data before using it for Artificial Intelligence and Machine Learning.

Data Science Connection:

Data Science uses statistical concepts to analyse data, identify patterns and obtain useful information that can support decision-making and AI applications.

1. Important Concepts in Statistics

Population and Sample

When studying data, it is important to understand the difference between a population and a sample.

Term Meaning School Example
Population The entire set of data or individuals being studied. Every student in India.
Sample A smaller part selected from the population for study. Students of one class.

Mean

The Mean is the average value of a set of numbers.

Mean = Sum of all values ÷ Number of values
Example:

Suppose five students score: 60, 70, 80, 90 and 100.

Mean = (60 + 70 + 80 + 90 + 100) ÷ 5
Mean = 400 ÷ 5 = 80

Therefore, the mean score is 80.

Median

The Median is the middle value when data is arranged in ascending or descending order.

Example:

Consider the values: 10, 20, 30, 40, 50.

The middle value is 30. Therefore, the median is 30.

Mode

The Mode is the value that occurs most frequently in a dataset.

Example:

Consider the marks: 40, 50, 50, 60, 70.

Since 50 occurs most frequently, the mode = 50.

Mean vs Median vs Mode

Concept Meaning Easy Way to Remember
Mean Average of all values. Add and divide.
Median Middle value after arranging data. Middle number.
Mode Most frequently occurring value. Most common number.

Distributions

A distribution describes how values in a dataset are spread or occur.

Distribution charts can help us understand patterns in statistical data.

Normal Distribution

A Normal Distribution is generally symmetrical and has a characteristic bell-shaped pattern.

Most values occur around the middle, while fewer values occur towards the extremes.

School Example:

If most students in a class score around the average marks and only a few students score extremely high or extremely low marks, the distribution may resemble a normal distribution.

Skewed Distribution

A Skewed Distribution occurs when the data is not symmetrical and values are concentrated more towards one side.

In simple terms, the distribution becomes lopsided.

School Example:

Suppose most students borrow 1–3 books every month, but a few students borrow a very large number of books. The distribution may become skewed because of these unusually high values.

Probability

Probability represents the likelihood or chance of an event happening.

AI Example:

An AI system may estimate the probability that a student will pass an examination based on factors such as study time, previous performance and attendance.

Variance

Variance measures how spread out the values in a dataset are from the average.

Remember:

Small variation generally means values are closer together, while greater variation means values are more spread out.

Outlier

An Outlier is a data point that is unusually far away from the other values in a dataset.

School Example:

Imagine that most students in a class score between 40 and 60 marks, but one student scores 100.

The score of 100 may be considered an outlier in that particular dataset.

Statistical Concepts – Quick Comparison

Concept What does it tell us?
Population The complete group being studied.
Sample A smaller part of the population.
Mean The average value.
Median The middle value.
Mode The most frequently occurring value.
Distribution How values are spread across a dataset.
Probability The likelihood of an event.
Variance How spread out the data is.
Outlier An unusually distant data value.

2. Orange Data Mining

Orange Data Mining is a free, open-source tool used for data mining, data visualisation and Machine Learning.

One of the important features of Orange is its visual, drag-and-drop interface.

Think of Orange as LEGO for Data Science:

Different components or widgets perform different tasks. These widgets can be connected together to create a complete data analysis or AI workflow.

The Canvas

The Canvas is the workspace in Orange where users place and connect different widgets.

It provides a visual representation of the complete AI or Machine Learning workflow.

Widgets

Widgets are individual components in Orange. Each widget performs a particular task.

Widget Purpose
File Used to load or upload a dataset.
Data Table Used to view data in rows and columns.
Select Columns Used to select appropriate features and labels from the dataset.
Scatter Plot Used to visualise relationships or patterns between variables.
Test and Score Used to evaluate how well a Machine Learning model performs.

Orange Workflow – Basic Idea

Data → Explore → Select Features → Build Model → Evaluate

The exact workflow depends on the problem and the dataset, but the visual workflow makes the process easier to understand.

3. AI Project Cycle in Orange Data Mining

Palmer Penguins Case Study

Real-World AI Case Study

Researchers collected information about penguins and wanted to identify their species using data rather than relying only on visual observation.

The penguins belonged to three species:

  • Adélie
  • Chinstrap
  • Gentoo

The objective is to use the available characteristics of a penguin to predict its species.

Problem Scoping

The first step is to understand the problem clearly.

Problem:

Can we use characteristics of a penguin to predict whether it belongs to the Adélie, Chinstrap or Gentoo species?

Data Acquisition and Exploration

The dataset contains different characteristics of penguins. These characteristics are called features.

Data Item Meaning Role
Island The island where the penguin lives. Feature
Culmen Length Length of the penguin's beak. Feature
Culmen Depth Depth of the penguin's beak. Feature
Flipper Length Length of the penguin's flipper. Feature
Body Mass Weight of the penguin. Feature
Species The species to be predicted. Target Label
Important Exam Point:

In the Palmer Penguins case study, Species is the Target Label, while characteristics such as body mass and flipper length are Features.

Mapping the AI Project Cycle to Orange

Stage What Happens Orange Component / Action
1. Data Acquisition Obtain the penguin dataset. Use the File Widget.
2. Data Exploration Examine the available data and identify useful information. Use Data Table and other visualisation tools.
3. Data Preparation Select appropriate features and the target. Use Select Columns.
4. Modelling Build a Machine Learning model. Connect a suitable model such as a Tree.
5. Evaluation Check how well the model performs. Use Test and Score.

Step-by-Step Palmer Penguins Workflow

Step 1 – Load Data: Use the File widget to load the penguin dataset.

Step 2 – Explore Data: Use the Data Table widget to inspect the dataset.

Step 3 – Select Data: Use Select Columns to identify appropriate features and the target label.

Step 4 – Build Model: Connect a Machine Learning model such as a Tree.

Step 5 – Evaluate: Use Test and Score to evaluate the model's performance.

Feature vs Target Label

Term Meaning Palmer Penguins Example
Feature An input characteristic used by the model. Island, Body Mass, Flipper Length, Culmen Length and Culmen Depth.
Target Label The output or result that the model attempts to predict. Species.

Real-Life Example – Library Outlier

Suppose the average student in a school library borrows 2 books per month.

One student borrows 50 books per month.

This unusually high value may be considered a statistical outlier.

The school could investigate the reason or recognise the student for exceptional reading habits.

Real-Life Example – Sports Day Predictor

Imagine a student develops a project in Orange to study the relationship between practice hours and running performance.

A regression model can be used to study the relationship between variables and make predictions.

y = mx + c

Here, the input variable can be represented by x and the predicted output by y.

Real-Life Example – Lunchroom Distribution

A teacher records the times at which students enter the school canteen during recess.

The data can be plotted as a distribution to identify patterns in student movement.

If the distribution is right-skewed, the data has a longer tail towards the higher values. Such analysis can help the school understand student movement and plan resources.

Case Study: The "Speeding Ticket" AI

Case Study

A student uses MS Excel and the Analysis ToolPak to study the relationship between the speed of a car and the distance required for the car to stop.

The student creates a Scatter Plot and observes a line represented by:

y = mx + c

Question 1

If y represents Stopping Distance and x represents Speed, which is the Independent Feature?

Answer:

Speed (x) is the independent feature.

The stopping distance depends on the speed of the car, so speed is treated as the independent variable in this relationship.

Question 2

If the R-squared value on the Excel chart is 0.99, what does this indicate?

Answer:

It indicates a very strong goodness of fit for the fitted regression relationship.

An R-squared value close to 1 means that the fitted line explains a very large proportion of the variation in the observed data.

Competency-Based Questions – 4.2

1. Multiple Choice Question

Q. Which widget in Orange Data Mining would a student use if they want to manually inspect the rows and columns of a dataset?

a) Scatter Plot
b) File
c) Data Table
d) Test and Score

Answer: c) Data Table

Explanation: The Data Table widget allows the user to inspect the data in rows and columns.

2. Assertion and Reasoning

Assertion (A): In the Palmer Penguins case study, Body Mass is considered a label.

Reason (R): A label is the special feature that represents the outcome the AI is trying to predict.

a) Both A and R are true and R is the correct explanation.
b) Both A and R are true but R is not the correct explanation.
c) A is false but R is true.
d) Both A and R are false.

Answer: c) A is false but R is true.

Explanation: Body Mass is an input feature, whereas Species is the target label that the model is trying to predict.

3. Application Task – Clustering

Scenario: A school has a dataset containing information about 1,000 students. The school wants to group students into categories such as:

  • Athletes
  • Artists
  • Coders

However, there are no pre-existing labels for the students.

Q. Which category of modelling should be considered?

Answer:

Unsupervised Learning / Clustering.

Since predefined labels are not available, the system can identify patterns and group similar students based on their characteristics.

4. Application-Based Question – Statistics

A teacher records the number of books borrowed by students every month. Most students borrow 1–3 books, while one student borrows 50 books.

Q. What statistical concept may describe the value 50?

Answer: Outlier.

The value is unusually far from the majority of the observations.

5. Application-Based Question – Orange

A student has received a dataset in a file and wants to load it into Orange before analysing it.

Q. Which widget should be used first?

Answer: File Widget.

The File widget is used to load the dataset into the Orange workflow.

Important Exam Points

Concept Remember This
No-Code AI Build AI solutions using visual tools without traditional programming.
Low-Code Uses visual tools but may require some coding.
High-Code Traditional programming with extensive manual coding.
Orange Open-source visual data mining and Machine Learning tool.
Population Complete group being studied.
Sample Smaller part of a population.
Mean Average value.
Median Middle value after arranging data.
Mode Most frequently occurring value.
Outlier An unusually distant data value.
Feature Input characteristic used by an AI model.
Target Label Output that the model attempts to predict.
Palmer Penguins Label Species.
Palmer Penguins Features Island, Culmen Length, Culmen Depth, Flipper Length and Body Mass.
File Widget Loads data.
Data Table Displays rows and columns of data.
Select Columns Selects features and target.
Scatter Plot Helps visualise relationships between variables.
Test and Score Evaluates Machine Learning models.

Unit 4 – Quick Revision

  • No-Code AI: Building AI solutions without writing traditional programming code.
  • Low-Code: Visual development with some coding when required.
  • High-Code: Traditional development involving manual programming.
  • Orange Data Mining: An open-source visual tool for data mining, visualisation and Machine Learning.
  • Teachable Machine: A tool that can be used to train models using images, sounds or poses.
  • Lobe AI: A simple tool for creating custom image classification models.
  • Population: Complete group being studied.
  • Sample: Smaller part of the population.
  • Mean: Average of the values.
  • Median: Middle value after arranging the data.
  • Mode: Most frequently occurring value.
  • Normal Distribution: Generally symmetrical and bell-shaped.
  • Skewed Distribution: A distribution that is lopsided towards one side.
  • Probability: Likelihood of an event occurring.
  • Variance: Measures how spread out data is from the average.
  • Outlier: An unusually distant data value.
  • Canvas: Workspace where Orange widgets are placed and connected.
  • Widget: An individual functional component in Orange.
  • Feature: Input characteristic used by the model.
  • Target Label: Output that the model attempts to predict.
  • Palmer Penguins Target: Species.
  • File Widget: Used to load data.
  • Data Table: Used to inspect data in rows and columns.
  • Select Columns: Used to select features and target variables.
  • Scatter Plot: Used to visualise relationships and patterns.
  • Test and Score: Used to evaluate Machine Learning models.
  • Clustering: Groups similar data when predefined labels are not available.
One-Line Memory Trick:

No-Code → Statistics → Orange → Widgets → Features → Target → Model → Evaluation