701 questions
Advice
1
vote
2
replies
54
views
Numerical meaning of Lightgbm Gain feature importance
I am using lightgbm to solve a binary classification problem. After training process, I use gain importance to analyse the model. The top 20 feature importance (gain) looks like the following
| ...
1
vote
1
answer
129
views
How to handle unstable best_iteration in LightGBM when using Optuna for hyperparameter optimization?
I'm using Optuna to optimize LightGBM hyperparameters, and I'm running into an issue with the variability of best_iteration across different random seeds.
Current Setup
I train multiple models with ...
- 177
0
votes
0
answers
86
views
tidymodels lightgbm hyperparameters training issue
I have an issue with training lightgbm models through tidymodels.
There seems to be some sort of issues in how the hyperparameters are translated between tidymodels and lightgbm.
This is my code:
...
- 1,089
1
vote
1
answer
60
views
Lightgbm early_stopping: min_delta doesn't work
I was using lightgbm with early_stopping and min_delta, but according to the result, min_delta seems to have no effect.
final_model = lgb.train(
params,
train_data,
...
- 65
0
votes
0
answers
92
views
How to use GPU to train LGBM model from SynapseML
I have gone through the entire docs on SynapseML LightGBMRanker module. There was no attribute I could pass to tell the model to use CUDA GPU for training.
For e.g. for the plain LightGBM library, ...
5
votes
1
answer
183
views
Darts and LightGBM: original column names cannot be retrieved for feature importance
Problem
I am running a LightGBMModel via Darts with some (future) covariates. I want to understand the relevance of the different (lagged) features.
In particular, I would like to retrieve the feature ...
6
votes
1
answer
708
views
Constructing custom loss function in lightgbm
I have a pandas dataframe that records the outcome of F1 races:
data = {
"Race_ID": [0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4, 4, 4],
"Racer_Number": [1, 2, 3, ...
- 592
1
vote
0
answers
35
views
Cannot reappearance mse in lightgbm custom objective when set feature_fraction
import numpy as np
import lightgbm as lgb
def custom_mse_objective(preds, train_data):
labels = train_data.get_label()
grad = (preds - labels)
hess = np.ones_like(labels)
return grad, ...
- 43
3
votes
1
answer
508
views
lightgbm force variables to be in splits
Im trying to find a way to train a lightgbm model forcing to have some features to be in the splits, i.e.: "to be in the feature importance", then the predictions are afected by these ...
- 488
1
vote
1
answer
215
views
What is max_evals in hyperopt fmin function?
I'm trying to run a multi-classification problem. I have run a baseline lightGBM model with around 80% accuracy rate. I'm trying to further fine-tuning its hyperparameter using Hyperopt. However, when ...
- 21
3
votes
1
answer
75
views
Reproduce LGBMRegressor predictions by manually aggregate the values
I am trying to reproduce by myself the LGBMRegressor predictions so when I succeed I will switch mean with median. But for now it seems that I am not able to.
Here is a simple script that I created ...
- 173
-1
votes
1
answer
111
views
Why lightgbm .predict function has probabilities not between 0 and 1? [closed]
I wanna understand why in this code, I get the following results:
# Import necessary libraries
import pandas as pd
from sklearn.metrics import f1_score
from sklearn.model_selection import ...
- 151
2
votes
1
answer
166
views
lightgbm.cv: cvbooster.best_iteration always returns -1
I am migrating from XGBoost to LightGBM (since I need it's exact handling of interaction constraints) and I am struggling to understand the result of LightGBM CV. In the example below, the minimum log-...
- 514
1
vote
2
answers
134
views
Why can't I wrap LGBM?
I'm using LGBM to forecast the relative change of a numerical quantity. I'm using the MSLE (Mean Squared Log Error) loss function to optimize my model and to get the correct scaling of errors. Since ...
- 113
1
vote
0
answers
118
views
Custom Labelling in Multi-Class Classification in XGBoost/ LightGBM
I have the following dataframe which records the IQ, Hours (number of hours of studying) and Score (past exam score for student 1,2,3,4 in different classes (Class_ID) and I would like to use these ...
- 592
