What is Data?
Collection of data objects and their attributes. |
Attribute is a property of an object |
Variable, field, characteristic, feature |
Collection of attributes decribes an object |
Record, point, case, sample, entity, instance, observation |
Type of attributes
Discrete |
Finite (countably) |
Integer |
Zip, Counts |
Continuous |
Real numbers |
Floating points |
Temp., height, weight |
Hierarchy of attributes types
Qualitative |
Nominal |
Category (=, !=) |
ID, zip, eye, color |
| |
Ordinal |
Ranked (>,<) |
Grades, {low, med., high} |
Quantitive |
Interval |
Distance (+,-) |
Dates, temp (C/F) |
| |
Ratio |
Zero means absence (*, /) |
Length, time, temp(K) |
|
|
Type of data sets
Record |
Collection of dataobjects and their attributes |
Table |
Relational |
Collection of data objects and their relation |
Graph |
Ordered |
Ordered collection of data objects |
Sequence |
Data quality
High quality |
Are fit for their intended use |
| |
Correctly represent the phenomena they correspond to |
Problems |
Noise |
| |
Outliers |
| |
Missing values |
|
|
Noise
Definition |
Unwanted pertubation to a signal |
| |
Unwanted data |
Reasons |
Limits in measurement accuracy |
| |
Intererence from other signals |
| |
Measurement of attributes not related to the data modeling task |
Handling |
Exclude noisy attributes |
| |
Remove noise by filtering |
| |
Include a model of noise |
Outliers
Definition |
Data objects which are significantly different from most others |
Reasons |
Measurement errors |
| |
Natural property of data |
Handling |
Identify & exclude outliers |
| |
Model the outliers |
Missing values
Definition |
No value is stored for an attribute in a data object |
Reasons |
Information is not collected |
People decline to give their age |
| |
Attribute is not applicable |
Annual income is not applicable to children |
Handling |
Eliminate data objects |
| |
Estimate missing values |
e.g. average |
| |
Ignore the missing value in analysis |
|
