|
|
Data Cleansing
| Description: |
To increase the quality of results Machine Learning
techniques yield, when applied to large datasets, a
step of inspecting the data and removing or correcting
corrupt or misleading parts should be performed first.
Typical problems are contradictory or incomplete
information. This will confuse learning algorithms,
for it is known that learning in the presence of
noise is much harder, than in the case of correct
information. Please refer to the case studies
for a more detailed discussion on data
cleansing.
|
|
|
