Why Data Cleaning?

~ "Garbage In, Garbage Out": Bad data will lead to bad results, plain and simple.
~ It's hard for computers to judge whether the data makes sense or not.
~ To get accurate results, you need to remove errors from you data which confuses the algorithms.
~ It's time-consuming process but important.

What are the causes?

• Input Errors
• Duplicates
• Mangled Data
• Malfunctioning Sensors
• Lack of Standardization

Identifying Problems

• Range Constraints
• Data-Type
• Compulsory Constraints
• Unique Constraints
• Cross Field Constraints

Data Cleaning Techniques

• Removing missing data
• Direct correction
• Normalization
• Syntax errors
• Data Imputation
• Spell Check
• Filter Unwanted Outliers
• Remove Irrelevant Values
• Fix structural errors