DMBI notes #3

Dated 24-1-17

2. Data Exploration

• Cosine similarity

• Data visualization

1. Pixel oriented visualization techniques.
2. Geometric projection visualization techniques.
3. Icon based visualization technique.

Dated 25-1-17

• Forms of data preprocessing

• Data clearing
• Data integration
• Data reduction
• Data transformation

• Need for data processing

1. Reason for incomplete data.
2. Reason for noisy data.

• Data preprocessing techniques

1. Data cleaning

• Methods for data cleaning

1. Missing values

1. Ignore the tuple.
2. Fill in the missing values manually.
3. Use a global constant value “Unknown”
4. Use attribute mean
5. Use attribute mean for all similarities.

2. Noisy data

• Methods to remove noise.

1. binning