Skip to main content
Data Science
Ben Reiniger's user avatar
Ben Reiniger's user avatar
Ben Reiniger's user avatar
Ben Reiniger
Moderator
  • Member for 6 years, 5 months
  • Last seen this week
Profile Activity
Stats
12,253
reputation
1.1m
reached
478
answers
0
questions
Communities
View all
About

math PhD (research in combinatorics) turned data scientist

pronouns he/him/his

Badges
View all badges
3
gold badges
19
silver badges
57
bronze badges
Top tags
View all tags
281
Score
163
Posts
34
Posts %
212
Score
124
Posts
26
Posts %
186
Score
109
Posts
23
Posts %
167
Score
68
Posts
14
Posts %
112
Score
63
Posts
13
Posts %
90
Score
59
Posts
12
Posts %
Top posts
View all questions and answers
answer
29
L1 & L2 Regularization in Light GBM
Aug 12, 2019
answer
16
What would I prefer - an over-fitted model or a less accurate model?
Jan 13, 2020
answer
13
Does scikit-learn have a forward selection/stepwise regression algorithm?
Dec 29, 2020
answer
12
Using a random forest, would a RandomForest performance be less if I drop the first or the last tree?
Oct 5, 2020
answer
10
Is There a Way to Re-Calibrate Predicted Probabilities After Using Class Weights?
Sep 9, 2019
answer
10
Why GridSearchCV returns nan?
Mar 27, 2021
answer
9
How do we standardize arrays with NaN?
Jul 31, 2019
answer
9
Why decision tree needs categorical variable to be encoded?
May 17, 2019
answer
9
Can Boosted Trees predict below the minimum value of the training label?
Jul 7, 2020
answer
8
What is the proper way to use early stopping with cross-validation?
May 17, 2020
Top Meta posts
3
8
7
Do we need both orange and orange3 tags?
6
On AI-generated answers
5
Why most of the questions in data science are unaccepted?
4
Are "improve my ML model" questions on-topic?
4
Should we add "category-theory" tag?
Top network posts
View all network posts
22
Why are there extra terms $-p_i+q_i$ in SciPy's implementation of Kullback-Leibler divergence?
18
Sklearn components in pipeline is not fitted even if the whole pipeline is?
17
Calibration curve of XGBoost for binary classification
16
SKLearn warning "valid feature names" in version 1.0
13
Is it really so bad to do SMOTE on the training set before crossvalidation?