CBSE Class 6 AI Chapter 3: Simple Pattern Recognition & Decision Making — Complete Teaching Guide

Chapter 3 of the Class 6 Artificial Intelligence unit is Simple Pattern Recognition & Decision Making, and CBSE allots it 05 hours out of the 20-hour AI block. The single most important thing about teaching it: the new skill this year is not spotting patterns — students have done that since Class 3 — it is turning a spotted pattern into a written rule, using that rule to decide something, and then hunting for the case where the rule falls apart.

This is the chapter where computational thinking and artificial intelligence finally shake hands. Everything a machine learning system does at scale is what your class will do on paper this week: look at examples, notice regularity, express the regularity as a rule, act on it, and correct the rule when it gets an answer wrong.

There is no coding here. No Python, no notebooks, no requirement to open any AI tool. A stack of paper slips, a blackboard and forty minutes will teach this better than any software, and I will argue that case throughout. For the full picture of where this chapter sits, see the Class 6 CT & AI syllabus and the wider 2026–27 curriculum overview.

What the syllabus actually says

The CBSE handbook lists Chapter 3 of the Class 6 AI block in a single table row:

Simple Pattern Recognition & Decision Making — Identifying patterns in data or daily routines; Making simple decisions based on observations — 05 hours

Two topics, five hours. That is deliberately spare, and it leaves the teaching decisions to you. Here is how I would map those two lines to outcomes you can actually observe in a classroom.

The last row is not printed in the handbook as a separate topic. I add it because a decision you never check is not a decision, it is a guess, and this is exactly the habit that separates a Class 6 AI lesson from a Class 4 maths pattern exercise. Where I recommend something beyond the handbook, I will say so plainly, as here.

The concepts, explained

What a pattern is — and what it is not

A pattern is a regularity you can state in words and use to predict. That definition does real work. If a child cannot state it, they have not found it. If they can state it but it does not predict, it is a description, not a pattern.

Class 6 students should comfortably handle four cases, and the fourth is the one schools skip.

• Repeating — the same unit comes round again. The school timetable, the weekly cycle of PT periods, the traffic signal sequence at the junction near the gate.
• Growing — each step changes by a rule. Savings in a piggy bank, the electricity bill climbing month on month from March to June.
• Symmetric — one half mirrors the other. A rangoli, a temple gopuram, a butterfly, the number 12321.
• No pattern — genuinely random. Yesterday’s lottery numbers. The order in which the class comes back from lunch.

Teach the fourth case on purpose. The most useful sentence a Class 6 student can learn to say this year is “I looked and there is no pattern here.” Machines that cannot say that produce nonsense confidently, and so do children.

Patterns in daily routines

Start here, not with numbers. Every student already lives inside a dozen patterns and has never been asked to name one.

The 7:40 school bus. Idli on Tuesday. The 5:50 local being packed and the 6:20 having seats. Cricket on the ground every evening except when it rains. The temple bell. The traffic on the flyover that is fine at 7:15 and hopeless at 8:15.

Ask the class for ten routines and write them on the board. Then ask the better question: which of these has broken in the last one month, and why? Answers will come fast — a strike, a holiday, an exam schedule, a wedding in the family, unexpected rain. Those breaks are the raw material for the second half of the chapter.

This connects directly to the computational thinking work students do in the pattern recognition teacher training module, where the same routines are used to build abstraction.

Patterns in data

Data is just observations written down in an organised way, which is precisely what Chapter 2 on basic data concepts established. Now we look for structure inside it.

Two moves matter at Class 6. First, arrange the data so the pattern can show itself — a table sorted by day, a chart with the days along the bottom. Second, describe the pattern in words before touching numbers: “as the temperature goes up, more cold drinks are sold.”

Then comes the honest part. A pattern is not a promise. When two things move together in a small sample, that is a starting point for a question, not proof of anything. Cold-drink sales and temperature rise together, and heat does drive thirst — but ice-cream sales and drowning deaths also rise together, and ice cream does not cause drowning. Both are driven by summer. Class 6 students can hold that idea if you give them the ice-cream example; they find it funny and they remember it.

From pattern to rule

This is the step that is genuinely new in Class 6. A pattern noticed in the head is worth little; a pattern written as a rule can be handed to someone else, argued with, and tested.

Insist on the if–then form, written on the board, in the students’ own words:

• If the month is between June and September, then carry an umbrella.
• If the SMS contains “you have won” and comes from an unknown number, then it is spam.
• If the mango’s skin is mostly yellow and it smells sweet at the stalk, then it is ripe.
• If the batsman keeps hitting through the covers, then place a fielder at cover.

Notice that three of those four rules have two conditions joined by “and”. Single-condition rules are where beginners start and where beginners get caught. Push a strong group to add a second condition and to say out loud what it buys them.

Making a decision from observations

A rule with several conditions gets clumsy in prose, so introduce the decision tree. Do not call it a machine learning model, do not mention algorithms — call it a question ladder if you like. The point is that each question splits the cases into smaller groups until the answer is obvious.

The umbrella tree below is the best classroom example in India because every child has been wrong about it. Copy it onto the blackboard and let students argue about the order of the questions. That argument — which question should we ask first? — is the real lesson, and it is the same question a decision tree algorithm answers automatically at a level students will meet in Class 9 and beyond.

Testing where a rule fails

Here is the mistake schools actually make with this chapter. The class writes a rule, everyone claps, the period ends. Nobody ever tests the rule against a case that breaks it. The chapter then teaches confidence rather than thinking, which is the opposite of what an AI unit should do.

Budget the last third of every activity for breaking the rule. Give it a name the class enjoys — “rule-breaker hunt” works. Award a point to any student who finds a real case where the class rule gives the wrong answer.

Then classify what happened. A rule can be wrong in two different ways, and they cost different amounts:

• The rule said carry the umbrella and it did not rain. You carried a wet-looking thing for no reason. Cost: mild irritation.
• The rule said leave it at home and it poured. Cost: a soaked school bag and a ruined notebook.

Those are not equal, and students grasp that immediately. This is the seed of a genuinely important idea they will meet later in Chapter 4 on AI ethics and digital responsibility: when an automatic system gets a decision wrong, someone pays, and the two kinds of wrong do not hurt the same person equally.

How to teach it — the five hours

Period 1. Twelve sequences on the board, three of each type from Figure 1, shuffled. Students sort them into four columns in their notebooks. The random ones will cause argument; that argument is the lesson. Close by writing the class definition of a pattern: something regular enough to describe in one sentence and use to predict what comes next.

Period 2. Move from shapes to life. Students list ten routines of their own, then mark which are daily, which weekly and which seasonal. Then the sharp question: name one that broke this month. Collect the breaks on the board. Keep that list — you will use it in Period 5.

Period 3. Bring a small table of real data. School attendance for the last two weeks works well and costs nothing. Students find the highest day, the lowest day, and the day that does not fit. Introduce the word outlier only if your class is comfortable; “the odd one out” is fine. Insist they propose a reason for the odd day. This links tightly to the data-handling work in Chapter 2 and to the maths chapter on data handling.

Period 4. The paper-slip classifier, described in full below. This is the period that carries the chapter. Do not substitute a slideshow.

Period 5. Rule-breaker hunt and repair. Every group hands its rule to another group, whose job is to break it. Groups then rewrite their own rule with one extra condition and note what the extra condition cost them in simplicity. End with Figure 5 on the board.

If your teachers are new to this style of teaching, the CT & AI teacher training programme runs these five periods as a live demonstration, and the school AI readiness score will tell you honestly whether your school is prepared for the 2026–27 rollout.

Unplugged activities

Activity 1 — The paper-slip classifier

Time: 35 minutes. Materials: 40 paper slips per group, each carrying three or four attributes of an object; one blank sheet per group for the rule; two labelled boxes or two chalk circles on the floor as bins.

Prepare the slips yourself before class. Use mangoes: each slip reads a colour (green / half yellow / yellow), a firmness (hard / slightly soft / soft), a smell (none / faint / sweet), and on the reverse, hidden, the true answer — RIPE or NOT RIPE. Make about a third of the slips awkward on purpose: yellow but hard, green but sweet-smelling.

1. Groups spread out their 40 slips, reverse side down. Nobody peeks.
2. Each group agrees on one question that will split the slips into two piles. Write the question on the sheet.
3. Sort all 40 slips by that question.
4. Agree on a second question for one or both piles. Sort again.
5. Write the finished rule in if–then form, for example: If the skin is yellow AND it smells sweet, then RIPE.
6. Place every slip into the RIPE bin or the NOT RIPE bin according to the rule.
7. Now turn the slips over. Count the mistakes. Write the number on the sheet in large figures.
8. Look only at the mistakes. What do they have in common? Add or change one condition and re-sort just those slips.

Discussion question: Which group had the fewest mistakes, and did they get there with the simplest rule or the most complicated one? Is a rule with six conditions still useful to a fruit seller in a hurry?

Activity 2 — Our own week of data

Time: 15 minutes of set-up, then five minutes a day for a week, then 30 minutes to analyse. Materials: one grid sheet per student; the class register; a wall chart.

1. The class agrees on one thing to count every day for five days. Good choices: number of students absent, number of students who brought an umbrella, minutes the class took to settle after lunch, number of vehicles passing the gate in two minutes at 8:00.
2. Assign two students per day as recorders. They write the number on the wall chart before the last bell.
3. On day six, students copy the five numbers into their notebooks and draw a simple bar chart.
4. Each student writes one sentence describing the pattern, and one sentence about the day that does not fit.
5. In pairs, students write an if–then rule predicting the number for next Monday.
6. The following Monday, actually measure it. Compare against every pair’s prediction. Nobody is marked for being wrong; everybody is asked why.

Discussion question: Five days is a very small sample. What would you believe more strongly if you had counted for five weeks instead? Name one thing that five weeks of data still would not tell you.

Activity 3 — Spam or not spam, by hand

Time: 25 minutes. Materials: 20 printed “SMS messages” on card, half genuine (bank OTP, school circular, courier delivery), half spam (lottery wins, free recharge, suspicious links). No phones needed and none should be used.

1. Groups read all 20 cards and sort them into SPAM and GENUINE using instinct alone. Record the piles.
2. Now the groups must justify themselves: list the words and features that made them say spam. Expect “won”, “free”, “click”, “urgent”, unknown sender, spelling mistakes, too many capital letters.
3. Write a rule using at most three of those features.
4. Re-sort all 20 cards using only the rule, ignoring instinct. Count where the rule disagreed with instinct.
5. The teacher reveals the true labels. Count mistakes.
6. Discuss the genuine message the rule wrongly flagged as spam. What did that cost the person?

Discussion question: Your phone’s spam filter uses a rule far more complicated than yours, and it still gets some wrong. Would you rather it blocked a little too much or a little too little? See also the shared vocabulary in the AI glossary.

Worksheet

Section A — Recall

1. Define a pattern in one sentence, in your own words.
2. Name the four kinds of pattern discussed in class and give one example of each.
3. What does an if–then rule look like? Write the general shape without using any real example.
4. What is an outlier? Why should we never simply delete one?

Section B — Apply

5. Complete the sequence: 3, 6, 11, 18, 27, ____, ____. State the rule you used.
6. Complete the sequence: square, circle, circle, square, circle, circle, ____, ____.
7. The electricity bill for a house was Rs 620, Rs 780, Rs 1,150, Rs 1,480 for March, April, May and June. Describe the pattern and give one reason for it.
8. A local train is crowded at 8:10 a.m. and empty at 11:30 a.m., every weekday. Write the rule in if–then form that would help a traveller choose a train.
9. Use the umbrella decision tree from class. It is the 12th of December, the sky is dark grey, and it rained yesterday. What does the tree decide?
10. Here are Monday-to-Friday absences: 4, 3, 4, 11, 3. Which day is the odd one out, and give two possible reasons.

Section C — Think

11. Find the case where this rule fails: If a fruit is yellow, then it is ripe. Give two different fruits that break it, and explain each.
12. Sales of ice cream and sales of umbrellas both rise in some months in Mumbai. Does one cause the other? Explain what is really going on.
13. Your rule for spotting spam SMS wrongly marked your school’s fee reminder as spam. Which word in your rule caused it? Rewrite the rule so it stops making that mistake without letting all spam through.
14. Look at these numbers: 8, 3, 15, 2, 9, 6. A classmate insists there is a hidden pattern. What would you say, and what would convince you they are right?
15. Design a two-question decision tree that decides whether the class should play cricket or stay indoors during the games period.

Answer key

1. A regularity you can describe in one sentence and use to predict what comes next.
2. Repeating (timetable), growing (savings), symmetric (rangoli), no pattern (lottery numbers). Accept any reasonable examples.
3. If [condition], then [decision or conclusion]. Conditions may be joined with AND or OR.
4. A value that does not fit the pattern of the rest. Never delete it, because it usually has a reason worth knowing.
5. 38, 51. Differences are 3, 5, 7, 9, 11, 13 — each gap grows by 2.
6. square, circle. The repeating unit is square-circle-circle.
7. The bill grows every month, and the jumps get bigger. Summer heat means more fan and cooler use.
8. If the time is a weekday morning rush hour, then expect a crowded train and travel earlier or later. Accept any sensible if–then form.
9. Not June to September, so the right branch; sky is dark grey, so YES; the tree decides CARRY IT. Yesterday’s rain is not asked on this branch.
10. Thursday, with 11 absences. Reasons might include a viral fever going around, heavy rain, a local festival or a bus breakdown.
11. A yellow lemon is not ripe in the eating sense; a yellow banana with green ends is not fully ripe; a ripe red apple is not yellow at all, so the rule also misses ripe fruit. Any two well-explained cases earn full credit.
12. No. Mumbai’s monsoon months are also hot and humid, so both go up together because of the season. Two things moving together does not mean one causes the other.
13. Likely “urgent” or “last date” or “pay now”. A better rule adds a sender condition: if the message contains a payment demand AND the sender is not in my contacts AND there is a link, then treat it as spam.
14. Say that no repeating, growing or symmetric pattern is present, so as far as we can tell there is none. What would convince me: the classmate states a rule that correctly predicts the next two numbers I have not shown them.
15. Any sensible tree, for example: Is it raining now? Yes to indoors. No to: Is the ground still wet from morning rain? Yes to indoors, no to cricket.

Common misconceptions

Assessment

At this stage the CT & AI curriculum is competency-based and largely internal. Assess the thinking, not the recall, and keep the evidence inside the classroom: notebooks, group sheets, the wall chart, a short oral question. There is no requirement I am aware of for a separate written examination in this subject at Class 6, and I would not invent one.

Four things are worth assessing.

• Can the student describe a pattern precisely enough for someone else to use it?
• Can the student write a workable if–then rule?
• Can the student find a case where a rule fails, including their own rule?
• Can the student explain the consequence of a wrong decision?

Record the level per student per criterion once, at the end of Period 5, from work you have already seen. Do not create a separate assessment event; it will eat an hour you do not have. Teachers building a portfolio approach across the year will find the sequencing useful in AI literacy for teachers.

Interdisciplinary links

Class 6 carries 40 hours of interdisciplinary projects alongside the 40 hours of computational thinking and 20 hours of AI. Chapter 3 is the single most portable chapter in the AI block, because pattern-and-rule thinking belongs everywhere.

• Mathematics. The NCERT Class 6 chapters Patterns in Mathematics, Number Play and Symmetry map onto this chapter almost line for line. Project: collect twenty number patterns from the maths textbook and classify each as repeating, growing or symmetric, then write the rule for each.
• Science. Project: a fortnight of weather observation — sky condition, temperature if a thermometer is available, whether it rained — then build a decision tree that predicts rain and score it against the next week.
• Social science. Project: patterns in a village or ward census table — population by age group, literacy by gender — with one written rule and one honest note about what the numbers cannot tell you.
• Languages. Project: rhyme scheme and metre in Hindi or English poems as repeating patterns; then write a rule that decides whether a new couplet fits the scheme.
• Art and craft. Project: design a rangoli or block-print motif with a stated symmetry rule, then swap rules with another student and reproduce their design from the rule alone. If the reproduction fails, the rule was incomplete.
• Physical education. Project: record where the ball goes for one batsman over twenty deliveries, find the pattern, and set a field. Then bowl to them and see if the field held.

The computational thinking half of the year supports all of this. See the companion modules on decomposition, abstract thinking and algorithmic thinking, and the overall CT and AI pillar guide.

Frequently asked questions

Do Class 6 students need to write code for this chapter?

No. There is no coding and no Python anywhere in the Class 6 AI block. Chapter 3 is conceptual and applied. Paper slips, a blackboard and a notebook are sufficient, and in my experience they produce better discussion than any software.

How is this different from the pattern work students did in Classes 3 to 5?

Earlier classes stop at spotting and extending a pattern. Class 6 adds three steps: writing the pattern as an if-then rule, using that rule to make a decision, and deliberately hunting for the case where the rule fails. The failure hunt is the genuinely new skill.

How many hours does CBSE allot to this chapter?

Five hours, as one of four chapters in the 20-hour introductory AI block for Class 6. The full Class 6 allocation is 100 hours: 40 hours of advanced computational thinking, 20 hours of AI and 40 hours of interdisciplinary projects.

Do we need a computer lab or an AI tool to teach it?

No. Nothing in Chapter 3 requires a device. If your school has a lab, use it for the interdisciplinary projects instead, where students can chart their collected data. The pattern-to-rule teaching itself is stronger unplugged.

What if my students insist there is a pattern when there is not one?

Ask them to state the rule, then test it on two values you have kept hidden. If the rule fails, the pattern was imagined. Doing this once in front of the whole class fixes the habit faster than any explanation.

How should this chapter be assessed?

Competency-based and largely internal. Assess whether a student can describe a pattern precisely, write a testable rule, find a case where a rule fails, and explain the consequence of a wrong decision. Use the four-criteria rubric in this guide and record it from work already produced in class.

PiyushAI builds classroom-ready material for the CBSE Computational Thinking & Artificial Intelligence curriculum, founded by Piyush Wairale (IIT Madras). Start with the CT and AI pillar guide, check where your school stands with the school AI readiness score, and bring your staff up to speed through CT and AI teacher training. Students can work through AI literacy for students alongside the syllabus.

Related: Class 6 AI Chapter 1: Introduction to AI and everyday examples · Chapter 2: Basic data concepts · Chapter 4: AI ethics and digital responsibility · Class 5 syllabus · Class 6 syllabus · Class 7 syllabus · Teacher training: pattern recognition · AI glossary

Source: CBSE Computational Thinking & Artificial Intelligence handbook, Class 6 (2026–27), cbseacademic.nic.in.