Bank Exams Exam  >  Bank Exams Notes  >  Tips & Tricks for Government Exams  >  Tips & Tricks: Classification

Classification Tips and Tricks for Government Exams

Introduction

Classification (often seen as "odd one out") is a reasoning task that asks you to group items by a shared property and then identify the item that does not share that property. In such questions you are given a set of words, numbers, letter groups, pairs or images; all but one share a clear relationship or category. Your objective is to detect that relationship quickly and select the item that breaks the pattern.

These questions test observation, vocabulary, general knowledge, pattern recognition, logical analysis and elimination skills. They are common in logical reasoning (LR) sections of government and banking exam papers where speed with accuracy is important.

Types of Classification Reasoning

Classification questions appear in a variety of forms. The common types are listed below with a short explanation of each.

  • Odd word based on general knowledge: A set of words where one word does not belong to the same semantic class (for example, three animals and one mineral).
  • Odd number: A set of numbers where one number differs by parity, primality, digit-sum, divisibility, square/cube relationships or another numeric property.
  • Odd numeral pair/group: Pairs or groups of numbers where one pair does not follow the same relation 
  • Odd letter group: Letter sequences or pairs where one item does not follow the same alphabetical pattern (such as alternate letters, equal gaps, reverse order).
  • Odd pair based on synonyms/antonyms: Word pairs in which three have synonym/antonym relations but one pair is unrelated or has the opposite relation.
  • Odd image-based pair: Images where three share a category (tools, fruits, vehicles) while one does not.

How to Solve Classification Questions - Tips & Tricks

Use the following practical steps and strategies when you practise or face classification questions in an exam.

  1. Scan quickly for an obvious category: Immediately check whether items are animals, countries, jobs, units, instruments, colours, etc. If a clear semantic category exists, test each item against it.
  2. Look for grammatical or morphological differences: Check singular/plural, parts of speech, prefixes/suffixes, tense or form differences that may make one item distinct.
  3. Check numeric properties: For numbers test parity (even/odd), primality, perfect squares/cubes, digit sum, factors, multiples, arithmetic operations between pair elements and other invariants.
  4. Examine letter patterns: See whether letters are in alphabetical order, alternate letters, equal skips, mirror/reverse order, or follow ASCII distance rules.
  5. Compare relationships in pairs: For pairwise items, establish the relation in each pair (synonym, unit-measure, country-capital, instrument-measured quantity) and find the pair that differs.
  6. Use elimination first: If you can identify three items that clearly follow one rule, the remaining item is the odd one out - this is usually faster than proving the relation for all items.
  7. Watch for tricky wording and antonyms: Options may include words that appear related but are antonyms or diagnostic opposites; check meaning precisely.
  8. Time management: Do easy classification questions first. If a question seems to require lengthy checking, mark and return later.
  9. Familiar categories to practise: Animals (diet), countries (continents), professions (healthcare vs. technical), units vs. instruments, geometric vs. arithmetic relations, prime/composite tests, square/cube checks, and common word-relationship patterns (synonym, antonym, part-whole).
  10. Maintain a systematic checklist while solving: semantic category → grammatical form → letter pattern → numeric pattern → pairwise relation. Proceed until one item fails all checks for the common rule.

Examples

Example 1: Choose the pair of words which are different from others in the group.
(a) Length : Extend
(b) Duration : Period 
(c) Time : Span 
(d) Expanse : Stretch
Ans:
(d)

Solution:
(a) Length : Extension → related but not exact synonyms (extension is the act/result of extending length)
(b) Duration : Period → clear synonyms
(c) Time : Span → clear synonyms
(d) Expanse : Stretch → close synonyms (both refer to a wide/extended area)

Options (b), (c), and (d) show direct synonym relationships. Option (a) shows a derivative/related concept, not a true synonym.

Example 2: In the following question, find the odd one from the given alternatives.
(a) Linen
(b) Cotton
(c) Iron 
(d) Silk
Ans:
(c)
Solution:
Linen, Cotton and Silk are materials used for making clothes (textile fibres). Iron is a metal and not a textile material. Therefore Iron is the odd one out.

Example 3: In the following question, find the odd one from the given alternatives.
(a) Netherlands
(b) Brunei
(d) Croatia 
(e) Serbia
Ans:
(b)
Solution:
The Netherlands, Croatia and Serbia are countries in Europe. Brunei is in Southeast Asia. Therefore Brunei is the odd one out.

Example 4: In the following question, find the odd one from the given alternatives.
(a) Electrician
(b) Paediatrician
(c) Oncologist 
(d) Cardiologist 
Ans:
(a)
Solution:
Paediatrician, Oncologist and Cardiologist are medical doctors (specialists). Electrician is a technical trade profession, not a medical specialty. Hence Electrician is the odd one out.

Example 5: Select the odd number pair from the given alternatives. 
(a) 13, 169
(b) 17, 269
(c) 14, 196
(d) 16, 256
Ans:
(b)

Solution:

Examine whether the second number is the square of the first number in each pair.

13, 169 →

\(13^2 = 169\)

17, 269 →

\(17^2 = 289\)
269 ≠  289

14, 196 →

\(14^2 = 196\)

16, 256 →

\(16^2 = 256\)

Therefore (b) is the odd pair because its second number is not the square of the first.

Example 6: Choose the word which is least like the other words in the group.
(a) Tiger
(b) Lion
(c) Leopard
(d) Cow
Ans:
(d)
Solution: Tiger, Lion and Leopard are carnivorous predators (big cats). Cow is herbivorous. Hence Cow is the odd one out.

Example 7: Choose the odd pair.
(a) Volume : Liter
(b) Time : Second
(c) Length : Meter 
(d) Pressure : Barometer 
Ans:
(d)

Solution: In (a), (b) and (c) the second term is a unit of measurement for the first term (liter measures volume; second measures time; meter measures length). A barometer is an instrument used to measure pressure, not a unit. Therefore (d) is different.

Example 8: Choose the number which is different from others in the group. 
(a) 13
(b) 17
(c) 23
(d) 27
Ans:
(d)
Solution: 13, 17 and 23 are prime numbers. 27 is composite (27 = 3 × 3 × 3). Therefore 27 is the odd number.

Example 9: Choose the number pair which is different from others in the group
(a) 14 - 7 
(b) 24 - 3 
(c) 37 - 5 
(d) 42 - 2
Ans:
(c)

Solution:
Check if the first number is a multiple of the second number in each pair.

14 is a multiple of 7.

24 is a multiple of 3.

37 is not a multiple of 5.

42 is a multiple of 2.

Therefore (c) is the odd pair.

Example 10: Choose the letter pair which is different from others in the group.
(a) BD
(b) IK
(c) PN
(d) SU
Ans:
(c)

Solution:

Observe the alphabetic gaps.

BD → B to D is +2 (B → C → D) so these are alternate letters in forward order.

IK → I to K is +2, forward order.

SU → S to U is +2, forward order.

PN → P to N is -2 (P → O → N), i.e., reverse order.

All pairs are two letters separated by one letter but PN is in reverse order. Hence PN is the odd pair.

 Common Traps

  • Practice regularly: Build speed with a mix of semantic, numeric and letter-pattern questions.
  • Master vocabulary and GK: Strong word knowledge and awareness of continents, currencies, units, professions and instruments help quick elimination.
  • Don't assume unnecessary complexity: Always check for the simplest semantic relation before looking for exotic relationships.
  • Be careful with near-synonyms and antonyms: Options may include words that look related; read meanings precisely.
  • Use scratch paper for numeric checks: Simple divisions, square checks and prime tests can be done quickly to remove wrong options.
  • Avoid overfitting a pattern: If three items clearly match one rule, prefer that rule rather than inventing a special exception that fits only two items.

Summary

Classification questions require pattern recognition across meaning, grammar, numbers and letters. Apply a short checklist (semantic → grammatical → numeric → alphabetical → pairwise), eliminate quickly, and practise frequently to build speed and accuracy. When a numeric relation is involved, check squares, primes, multiples and digit-sum properties. For words, test synonyms, antonyms and category membership. With systematic practice these questions become fast and reliable scoring opportunities.

The document Tips & Tricks: Classification is a part of the Bank Exams Course Tips & Tricks for Government Exams.
All you need of Bank Exams at this link: Bank Exams

FAQs on Tips & Tricks: Classification

1. What is classification in machine learning?
Ans. Classification in machine learning is a technique used to categorize data into different classes or groups based on their characteristics or features. It involves training a model using labeled data to predict the class of new, unseen data points.
2. How does classification work in machine learning?
Ans. Classification in machine learning works by training a model on a labeled dataset, where each data point is associated with a class label. The model learns patterns and relationships between the input features and the corresponding class labels. Once trained, the model can be used to predict the class of new, unseen data points based on their features.
3. What are some popular algorithms used for classification in machine learning?
Ans. Some popular algorithms used for classification in machine learning include logistic regression, decision trees, random forests, support vector machines (SVM), and naive Bayes. Each algorithm has its own strengths and weaknesses, and the choice of algorithm depends on the specific problem and dataset.
4. What are the evaluation metrics used for assessing the performance of classification models?
Ans. There are several evaluation metrics used for assessing the performance of classification models. Some commonly used metrics include accuracy, precision, recall, F1 score, and area under the receiver operating characteristic curve (ROC AUC). These metrics provide insights into the model's ability to correctly classify different classes and handle imbalanced datasets.
5. Can classification models be used for text classification tasks?
Ans. Yes, classification models can be used for text classification tasks. Text classification involves assigning predefined categories or classes to text documents based on their content. Techniques such as bag-of-words, word embeddings, and recurrent neural networks can be used to represent and classify text data. Classification models have been successfully applied in various text classification tasks, including sentiment analysis, spam detection, and topic categorization.
Explore Courses for Bank Exams exam
Get EduRev Notes directly in your Google search
Related Searches
Tips & Tricks: Classification, Viva Questions, Tips & Tricks: Classification, Previous Year Questions with Solutions, Free, pdf , Semester Notes, video lectures, practice quizzes, ppt, Exam, mock tests for examination, MCQs, Tips & Tricks: Classification, shortcuts and tricks, Extra Questions, past year papers, Sample Paper, Important questions, Summary, Objective type Questions, study material;