Icon

GoGoX_​Cancellation_​Prediction

Data Understanding & Cleaning

Temporal Feature Engineering

Route & Pricing Feature Engineering

Model Preparation & Validation

EDA

EDA

Geographic Feature Engineering

MODEL PART 1

Model Test and Training

Model Interpretation

Unmatched Risk Scores

Column Renamer
Math Formula
Color Palette Designer
Color Palette Designer
Math Formula
Rule-based Row Filter
GroupBy
Column Renamer
Rule Engine
Math Formula
Rule Engine
Math Formula
Column Filter
Rule-based Row Filter
Sorter
Color Palette Designer
Booking Lead Time
Bar Chart
Excel Reader
Table View
Statistics
clean missing values
Missing Value
status
Value Counter
order vehicle
Value Counter
platform
Value Counter
requested vehicle type
Bar Chart
Sorter
Color Palette Designer
Waypoint band
Bar Chart
Column Renamer
Column Filter
Column Renamer
Column Filter
Duplicate Row Filter
combined pickup and dropoff into one table
Concatenate
String Manipulation
Classifies scheduled pickups into weekdays and weekends
Rule Engine
pickup time
Heatmap
target distribution
Pie Chart
payment method
Value Counter
GroupBy
changed name of nus to what the extension will understand
Rule Engine
driver vehicle type
Value Counter
Column Renamer
Geocoding
pickup location
Scatter Plot
GeoFile Reader
drop off location
Scatter Plot
Checked for any missing data
Statistics
verify that regions are correct
Value Counter
Spatial Join
GroupBy
Column Filter
check for missing info
Row Filter
pickup zone
Joiner
Column Renamer
remove active
Rule-based Row Filter
secured driver
Rule Engine
dropoff zone
Joiner
lead time classification
Rule Engine
Column Renamer
request hour
Date&Time Part Extractor
Column Renamer
lead time
Date&Time Difference
Rule-based Row Filter
booking context
Rule Engine
extracting out the hour
Date&Time Part Extractor
extracting out the day
Date&Time Part Extractor
Column Filter
request day
Rule Engine
extracting out the day
Date&Time Part Extractor
pickup zone
String Manipulation
dropoff zone
String Manipulation
classification of period of the day
Rule Engine
remove columns not needed for fair predictions
Column Filter
Statistics
waypoint count
String Manipulation
80 | 20 split
Table Partitioner
Base order price
Math Formula
Rule Engine
Random Forest Learner
75 | 35 split
Table Partitioner
model lead time
Rule Engine
AAUC = 0.726
ROC Curve
Rule-based Row Filter
Used for training
Random Forest Predictor
Rule Engine
Rule Engine
Math Formula
Column Renamer
Scorer
Math Formula
Sorter
Determine 0.16 is the best-fit threshold for a balance of accuracy and precision
Scorer
GroupBy
Trail and error for best-fit threshold
Rule Engine
Used to test the reserved 20% to evaluate predictive performance on unseen order
Random Forest Predictor
AUC = 0.726
ROC Curve
RowID
Applies the validation-selected 0.16 threshold to unseen test orders
Rule Engine
Rule-based Row Filter
AUC = 0.730
ROC Curve
Ranks booking features by their frequency of selection at the Random Forest's root splits
Sorter
Measures final classification performance
Scorer
Examines predictor usage across the first three tree levels to identify frequently selected booking characteristics
Table View
Rule-based Row Filter
Math Formula
RowID
GroupBy
Column Renamer
Column Renamer
Validated the generated risk scores against original predictions and order characteristics before business profiling
Table View
rename columns
Column Renamer
Math Formula
Converts Random Forest Unmatched prediction scores into a 0–100 risk index
Math Formula
Rule Engine
GroupBy
Checks the distribution, range and completeness of unmatched risk scores to identify potential anomalies
Statistics
Box Plot
Duplicate Row Filter
create target status
Rule Engine
Rule Engine
target_status
Value Counter
Number to String
String to Date&Time

Nodes

Extensions

Links

Download

To use this workflow in KNIME, download it from the below URL and open it in KNIME:

Download Workflow
Created by: hazel
Created at: 2026-10-06
On NodePit since: 2026-10-07
Last update: 2026-10-13
Created with KNIME version: v5.12.0

Deploy, schedule, execute, and monitor your KNIME workflows locally, in the cloud or on-premises – with our brand new NodePit Runner.

Try NodePit Runner!