Timeline for Why doesn't LightGBM perform better than SVC (linear kernel)? [sentiment analysis]
Current License: CC BY-SA 3.0
8 events
when toggle format | what | by | license | comment | |
---|---|---|---|---|---|
Jul 14, 2017 at 15:17 | answer | added | Dhruv Mahajan | timeline score: 2 | |
Jul 14, 2017 at 10:23 | comment | added | tuomastik | Try lowering the learning rate to 0.01 and increase the number of estimators to 3000. Be sure to include early stopping to not to overfit. Also, plot and study learning curves to understand whether to increase the number of estimators even more. | |
Jul 14, 2017 at 9:15 | comment | added | tktktk0711 | I found that the parameter( just my parameter: max_depth=3, learning_rate=0.1, n_estimators=1000) have significantly influence on error when I used it for regression. Sorry I haven't used the mode for classifier | |
Jul 14, 2017 at 9:11 | comment | added | Satjapong Meeklai | @tktktk0711Yes, I did grid search on LightGBM model. The result was 0.1 learning rate, 150 num_tree and 100 num_leaves. The acc of LightGBM above is the result of the model with these parameters | |
Jul 14, 2017 at 9:08 | history | edited | Satjapong Meeklai | CC BY-SA 3.0 |
deleted 2 characters in body
|
Jul 14, 2017 at 9:08 | comment | added | tktktk0711 | have you set different parameter of lightGBM | |
Jul 14, 2017 at 9:07 | history | edited | Satjapong Meeklai | CC BY-SA 3.0 |
deleted 2 characters in body
|
Jul 14, 2017 at 9:02 | history | asked | Satjapong Meeklai | CC BY-SA 3.0 |