Timeline for Classification model accuracy, roc auc score, f1 score 100%
Current License: CC BY-SA 4.0
15 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Oct 6, 2019 at 22:43 | comment | added | Zen | Would you mind posting a subset of your dataset? | |
| Oct 6, 2019 at 14:23 | comment | added | roundsquare |
I wonder if your features actually completely determine the value of the target. You have 14 features and only 6,825 records. Maybe check this by doing df.groupby(features).agg({'target: 'nunique'}) and see if any of the resulting values are greater than 1.
|
|
| Oct 6, 2019 at 14:19 | comment | added | Zen | What does the response/target variable account for? | |
| Oct 6, 2019 at 14:00 | history | bumped | CommunityBot | This question has answers that may be good or bad; the system has marked it active so that they can be reviewed. | |
| Oct 3, 2018 at 15:18 | comment | added | kbrose |
But assigning it to a new variable does remove it from the assigned variable features. Is there a single variable that you can remove which causes performance to drop?
|
|
| Oct 3, 2018 at 15:01 | answer | added | Jon Nordby | timeline score: 2 | |
| Oct 3, 2018 at 14:58 | comment | added | Jon Nordby |
That drop() call does not remove the targets column from df, since you did not specify inplace=True...
|
|
| Oct 3, 2018 at 12:43 | history | edited | sergio | CC BY-SA 4.0 |
added 744 characters in body
|
| Oct 3, 2018 at 12:40 | comment | added | sergio | I've added the features i'm using in the question above. I find difficult to believe that the good results are accurate, if I remove a lot of features I still get 100%, only if I have 2/3 features it drops to 50% | |
| Oct 3, 2018 at 11:59 | comment | added | Calimo | Without seeing the data it's going to be hard to answer. What makes you think you don't have one really good feature? | |
| Oct 3, 2018 at 9:31 | comment | added | sergio | I tried max_depth = 5 and got the same results. Before splitting I dropped the targets: features = df.drop(['target'], axis=1) | |
| Oct 3, 2018 at 9:27 | comment | added | Jakub Bartczuk | Are you sure you didn't leak target into features? It often caues such problems. Also did you try decreasing random forest's max_depth? | |
| Oct 3, 2018 at 6:45 | comment | added | SmallChess | I had a brief look but failed to find anything. Maybe your data was quite simple? | |
| Oct 3, 2018 at 6:40 | review | First posts | |||
| Oct 3, 2018 at 8:04 | |||||
| Oct 3, 2018 at 6:37 | history | asked | sergio | CC BY-SA 4.0 |