---
title: "Short note: Data preprocessing"
subject: "Data Mining"
chapter: 2
key: "semester_7/data_mining:data-preprocessing-note"
source: https://www.bctengineer.com/q/semester_7/data_mining/data-preprocessing-note
site: BCT Engineering
---

# Short note: Data preprocessing

Data Mining · Chapter 2: Data Pre-processing

- Asked in: year not given, 3 marks; 2063, 3 marks
- Video: [Lec-32: What is Data Preprocessing & Data Cleaning | Various Techniques with Example](https://www.youtube.com/watch?v=tDu_KIlXaB0) — Gate Smashers
- Answer checked by Claude Opus 5.5 (high effort) on 2026-10-14
- Page: https://www.bctengineer.com/q/semester_7/data_mining/data-preprocessing-note
- Whole chapter: https://www.bctengineer.com/md/semester_7/data_mining/chapter-2

## Question

Write a short note on Data Preprocessing.

## Short answer

Data preprocessing converts raw data into clean form for mining, addressing incompleteness, noise, inconsistency. Major tasks: cleaning (missing values, noise, outliers), integration (merge sources), transformation (normalization, aggregation), reduction (dimensionality, compression), discretization (numeric to intervals). Takes 60-80% of project effort.

## Answer

**Data preprocessing** is the set of techniques used to convert raw data into a clean and suitable form before data mining. Real data is incomplete (missing values), noisy (errors, outliers) and inconsistent (different codes, units), and low-quality data gives low-quality mining results.

**Major tasks**
1. **Data cleaning**: fill missing values (ignore tuple, mean, most probable value), smooth noise (binning, regression, clustering), remove outliers, fix inconsistencies.
2. **Data integration**: combine multiple sources; handle schema and entity matching, redundancy (correlation) and value conflicts.
3. **Data transformation**: normalisation (min-max, z-score, decimal scaling), aggregation, generalisation, attribute construction.
4. **Data reduction**: attribute subset selection, PCA, sampling, compression, so data is smaller but gives similar results.
5. **Data discretisation**: convert numeric attributes into intervals or concepts using binning or histograms.

Preprocessing often takes 60 to 80% of the total effort in a data mining project.

## Check yourself

The answer's key points as multiple-choice questions, with the correct option and why.

1. Why is preprocessing needed?
   A. Real data is incomplete, noisy and inconsistent
   B. Data is always clean
   C. To increase noise
   D. To skip mining
   Answer: A. Low-quality data gives low-quality results. (Key point: Why preprocess)

2. Which is a data transformation technique?
   A. Min-max, z-score or decimal scaling normalisation
   B. PCA
   C. Sampling
   D. Binning for missing values
   Answer: A. PCA and sampling are data reduction. (Key point: Transformation techniques)

3. Data reduction aims to:
   A. Make data smaller while giving similar results
   B. Add attributes
   C. Increase size
   D. Remove users
   Answer: A. Methods include attribute subset selection, PCA, sampling, compression. (Key point: Data reduction)

## Related in Data Mining

- [Data preprocessing and handling noisy data](https://www.bctengineer.com/question.md?key=semester_7%2Fdata_mining%3Apreprocessing-noisy-data) — Data Mining, chapter 2
- [Processes in data preprocessing](https://www.bctengineer.com/question.md?key=semester_7%2Fdata_mining%3Apreprocessing-processes) — Data Mining, chapter 2
- [Preprocessing steps, binning, normalization](https://www.bctengineer.com/question.md?key=semester_7%2Fdata_mining%3Apreprocessing-binning-normalization) — Data Mining, chapter 2