Icon

KNIME_​finalproject

Filter down to columns I am reviewing (update for final models)

Data Homogenization (Rows)

Constructing the Dependent Variable

Make DRDISTRACT into string, need to decide what to do with unknown values

Ejection cleaning - imputed unknown (7-9) to 0 ("No") as it is unlikely the person ejected if unknown and zero is mode of variable

Fire_Exp - no missing values, only 0 = No and 1 = Yes

Only Class 1 (6k or less) and Class 2 (6k-10k) are in our data

Aggregated 5:00-17:00 as "Day" and 17:01-04:59 as "Night" - also some unknowns, that I aggregated as "Night" since more common

Binned into "Lighted" and "Not fully lit" - dawn and dusk are consider the latter, same with Dark - Unknown lighting

Interstate highway has no unknown values, so just binning into Yes or No

Data Explorer
Row Filter
Numeric Binner
Row Filter
Numeric Binner
Missing Value
Reading & Joining Files
Data Explorer
Missing Value
Column Filter
Row Filter
Rule Engine
Numeric Binner
Missing Value
Numeric Binner
Rule Engine
Missing Value
Rule Engine
Rule Engine

Nodes

Extensions

Links