Data reduction by sampling

Data reduction by sampling

1. TYPES OF SAMPLES

   • Stratified sampling
     • Select samples independently within non-overlapping groups of the population, or strata.
       • For example,
         • to ensure that men and women are sampled in equal proportion.
   • Systematic or 1-in-n sampling
     • For non-random (structured) samples
     • When selection at random is difficult to obtain,
       • units can be sampled systematically (at a fixed interval) or sequentially.

2. When the dataset is extremely large, what is the issue?

   • , it may take a long time (can be up to days) for a computer to analyse the data and generate a meaningful model.
     • To reduce time in analysis,solution is?
       • a smaller set of data records can be randomly selected from the data source.
         • Thus, is known as random sampling
           • This random sample represents the original data population but is a lot smaller in data size.
             • With the smaller sample, data mining models can be generated more quickly to facilitate analysis.