www.flow.help.3_1_4.html Maven / Gradle / Ivy
Scoring history: GBM, DL Represents the error rate of the model as it is built. Typically, the error rate will be higher at the beginning (the left side of the graph) then decrease as the model building completes and accuracy improves. Can include mean squared error (MSE) and deviance.
Variable importances: GBM, DL Represents the statistical significance of each variable in the data in terms of its affect on the model. Variables are listed in order of most to least importance. The percentage values represent the percentage of importance across all variables, scaled to 100%. The method of computing each variable’s importance depends on the algorithm. To view the scaled importance value of a variable, use your mouse to hover over the bar representing the variable.
Confusion Matrix: DL Table depicting performance of algorithm in terms of false positives, false negatives, true positives, and true negatives. The actual results display in the columns and the predictions display in the rows; correct predictions are highlighted in yellow. In the example below, 0 was predicted correctly 902 times, while 8 was predicted correctly 822 times and 0 was predicted as 4 once.
ROC Curve: DL, GLM, DRF Graph representing the ratio of true positives to false positives. To view a specific threshold, select a value from the drop-down Threshold list. To view any of the following details, select it from the drop-down Criterion list:
- Max f1
- Max f2
- Max f0point5
- Max accuracy
- Max precision
- Max absolute MCC (the threshold that maximizes the absolute Matthew’s Correlation Coefficient)
- Max min per class accuracy
The lower-left side of the graph represents less tolerance for false positives while the upper-right represents more tolerance for false positives. Ideally, a highly accurate ROC resembles the following example.
Hit Ratio: GBM, DRF, NaiveBayes, DL, GLM (Multinomial Classification only) Table representing the number of times that the prediction was correct out of the total number of predictions.
Standardized Coefficient Magnitudes GLM Bar chart representing the relationship of a specific feature to the response variable. Coefficients can be positive (orange) or negative (blue). A positive coefficient indicates a positive relationship between the feature and the response, where an increase in the feature corresponds with an increase in the response, while a negative coefficient represents a negative relationship between the feature and the response where an increase in the feature corresponds with a decrease in the response (or vice versa).
To learn how to make predictions, continue to the next section.