You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi,
in chapter 9.6.6 a SHAP summary plot for a random forest model trained on the cervical cancer dataset is given. However, from what I checked, and from a previous explanation on chapter 9.5.2, the use of a hormonal contraceptive is correlated positively with an increase in cervical cancer risk. If the Hormonal Contraceptive feature (a yes/no categorical feature) is encoded by a 1 for a yes and a 0 for a no, then high values (red-coloured) of this feature should be seen on the x-axis of the summary plot as having a positive SHAP value and not a negative one as shown currently. Is my interpretation of this issue erroneous or is there a mistake in the plot?
Thank you for reading.
The text was updated successfully, but these errors were encountered:
FWijanto
changed the title
Colors of Hormonal Contraceptives feature from SHAP summary plot seems wrong.
Colors of Hormonal Contraceptives feature from SHAP summary plot seem wrong.
Nov 28, 2022
Hi,
in chapter 9.6.6 a SHAP summary plot for a random forest model trained on the cervical cancer dataset is given. However, from what I checked, and from a previous explanation on chapter 9.5.2, the use of a hormonal contraceptive is correlated positively with an increase in cervical cancer risk. If the Hormonal Contraceptive feature (a yes/no categorical feature) is encoded by a 1 for a yes and a 0 for a no, then high values (red-coloured) of this feature should be seen on the x-axis of the summary plot as having a positive SHAP value and not a negative one as shown currently. Is my interpretation of this issue erroneous or is there a mistake in the plot?
Thank you for reading.
The text was updated successfully, but these errors were encountered: