Data Preprocessing - Things to consider.
#prashraghav
sData in real world is inconsistent, noisy and have missing values. In order to improve the quality of data, preprocessing of data is required. There are several ways to preproces the data : data cleaning, data integration, data reduction or data transformation. Raw Data : Noisy : Data can have outliers which can be termed as values which do not follow the regular data pattern and are different from the rest. Data Can have anomaly and can have errors. Missing Data: data can have missing values. Inconsistent : Containing discrepancies in names or values. Aggregated data: Dataset contains only aggregated data. Sample Dataset :
Sample dataset is a US census -Income data and is taken from UCI ML Repsoitory.
Read More....
PROCESS EMAILS DAILY CLICK HER FOR MORE INFO.
