Questions tagged [catboost]
CatBoost is an open-source gradient boosting on decision trees library with categorical features support out of the box for Python & R
33
questions
1
vote
0
answers
21
views
Is there a way to set an exposure for a Poisson loss in Catboost?
I'd like to use Catboost for actuarial models (eg claims frequency). Although I see that Poisson loss is an option, I don't see that exposures are directly supported. How do people deal with this?
- 1,915
0
votes
0
answers
31
views
Custom loss function optimized for correlation for gradient boosted trees algorithms like XGBRegressor or LGBMRegressor
I have a rather classical tabular data prediction problem and more or less successfully using XGBRegressor or LGBMRegressor both using MSE as their loss function.
A slight deviation from the standard ...
2
votes
0
answers
58
views
Can I use multiple quantile regression to estimate the probability a dependant variable is above / below a certain value?
Let's say I have a dataset of characteristics of newly launched products in a retail environment, and the dependant variable Y is total $ sales in the first year of ...
- 267
1
vote
0
answers
35
views
Is there a better way to self-train on tabular data?
Context:
I'm training a classifier on some fraud data. Only a chunk of data is labeled (~2000) so I'm trying a self-training approach, what I'm doing for now is:
Iteratively training a model then ...
2
votes
1
answer
218
views
Why doesn't CatBoost Encoding cause target leakage?
I'm currently working on a fraud detection problem with a dataset of 300,000 rows and 500 columns, 70 of which are categorical with over 10 categories each. I'm facing memory constraints and exploring ...
- 451
0
votes
0
answers
59
views
Creating Brier Score loss function for Catboost in R
Catboost allows the use of Brier Score as a metric, but not for use as a loss function in model training.
I'm attempting to implement Brier score as a custom loss function in R, but I'm finding it a ...
- 21
0
votes
0
answers
412
views
Improving an underfit model with Catboost for regression problem
I'm currently using CatboostRegressor(iterations=500, random_seed=123, cat_features=['month_number', 'day_of_week', 'year']) for developing a 1-year predictive ...
- 143
1
vote
0
answers
493
views
Why are my catboost predictions pushed towards 0 and 1 and how to smooth it? [closed]
I am using catboost in a classification problem with extreme unbalanced classes (positive class consists of ~ 0.25% of all observations).
I fitted a model using catboost in R and got a very good ...
- 11
1
vote
0
answers
172
views
Total number of learned weights/parameters in CatBoost model?
I wasn't really able to find this anywhere in the documentation, or after searching through the model object itself (though it must be there somewhere, of course. The ...
- 111
1
vote
1
answer
287
views
Generalization of model performance (AUC) and tuning of a catboost classifier
I was wondering if it is good practice to overfit on the training data while tuning a catboost classifier for a binary outcome.
Wouldn't it be better to reguralize until validation error equals ...
- 155
3
votes
2
answers
665
views
How do Ordered Target Statistics work for CatBoost?
This question follows closely this paper .
I'm trying to fully understand how Ordered Target Statistics (TS) (for CatBoost) works. E.g. the CatBoost algorithm uses this method to group categorical ...
- 90
8
votes
1
answer
2k
views
Any reasons to prefer neural networks over boosting methods in tabular data?
Based on Kaggle winners data, it seems that ensemble boosting methods like XGBOOST, LIGHTGBM, CATBOOST are the top choices when dealing with structured or tabular data for maximizing the prediction ...
- 188
0
votes
0
answers
249
views
High precision and low recall but with a balanced dataset
When i evaluated my model (CatBoost classifier), I noticed that my model has high precision and low recall (Recall: 0.59, Precision: 0.89) but the classes are perfectly balanced (1: 45.5, 2: 54.5) and ...
1
vote
0
answers
99
views
Analytical expression of a CatBoost regression model in R
When adjusting a multiparametric regression model, an analytical expression that characterizes the fitted model (e.g., in a linear multiparametric regression, the equation is $\hat\beta= \hat\beta_o+\...
- 97
1
vote
1
answer
95
views
If evaluation set is the same as training set, why would the evaluation error be different from training error?
I understand the use of evaluation set for parameter tuning and over-fitting in general. The examples in the evaluation set should be unseen and different from training set.
However, in the following ...
- 11
1
vote
0
answers
250
views
Role of weighted quantile sketch in XGBoost, LightGBM and CatBoost
Have i understood it correctly if i say that "weighted quantile sketch" allows XGBoost to use histogram search on feature values for split finding?
Also, do LightGBM or CatBoost use ...
- 85
2
votes
0
answers
82
views
Looking for information regarding tree-based gradient boosting algorithms comparative performance on data sets with different underlying properties
I have a difficult time finding any theoretical or empirical comparative research regarding the tree-based gradient boosting algorithms on data sets with different underlying properties. Is there any ...
- 85
2
votes
1
answer
2k
views
Why does LightGBM Classifier gives some folks a probability of 1 of belonging in a class with log-loss?
I'm trying to use the LightGBM package in python for a multi-class classification problem and I'm baffled by its results.
For a minority of the population, LightGBM predicts a probability of 1 (...
- 774
2
votes
1
answer
2k
views
catboost does not overfit - how is that possible?
I'm fitting and evaluating a CatBoostRegressor and a XGBRegressor to the same regression problem. I tried matching their ...
- 3,220
2
votes
0
answers
323
views
Gradient boosting (GB) splitting methods (categorical features)
Regarding categorical features - ordinary trees treat categorical features in two main ways, CART - considers only binary splitting, those computing the mean response value (y_mean_i per each category ...
- 21
2
votes
1
answer
7k
views
L2 Regularization in CatBoost
I am studying the CatBoost paper https://arxiv.org/pdf/1706.09516.pdf (particularly Function BuildTree in page 16), and noticed that it did not mention regularization.
In particular, split selection ...
1
vote
1
answer
1k
views
negative value of prediction from a training set that only contains positives
a catboost model recently gave me a % of predictions that are negative while none of the training set contains negative values for the label.
How is this possible especially in cases where the % of ...
- 13
1
vote
1
answer
668
views
Model Size of Random Forest VS CatBoost
I trained models based on the same dataset, using random forest (sklearn) and CatBoost.
I use n_estimators=1000 for random forest, and n_estimators(iterations)=1000 for CatBoost. The random forest ...
- 15
0
votes
0
answers
27
views
Tweak Model or Process to Focus More on Category Values for Regression
I have built a machine learning model using Catboost and I noticed something as I examine the results. When I am looking at the results there is a big fluctuation in the errors between my test set and ...
- 11
3
votes
1
answer
950
views
Negative Feature Importance Value in CatBoost LossFunctionChange
I am using CatBoost for ranking task. I am using QueryRMSE as my loss function. I notice for some features, the feature importance values are negative and I don't know how to interpret them.
It says ...
- 97
1
vote
2
answers
735
views
AUPRC vs AUROC and updating training set in quasi-classification problem
I have an unbalanced classification problem (95% "0", 5% "1") regarding quality control."0" means "no problem" and "1" means "problem".
I'm not predicting real cases one by one, this is, my client ...
3
votes
1
answer
2k
views
Feature Interaction Strength in Catboost
I was wondering if anyone knew how the feature interaction strength is calculated in the catboost package. The documentation https://catboost.ai/docs/concepts/output-data_feature-analysis_feature-...
- 33
3
votes
1
answer
3k
views
(Low cardinality) categorical features handling in gradient boosting libraries
In some popular gradient boosting libraries (lgb, catboost), they all seems like can handle categorical inputs by just specifying the column names of the categorical features, and pass it into a ...
- 377
0
votes
1
answer
660
views
Catboost Regression. Function Extrapolation
I'm new at ML and have a problem with catboost. So, I want to predict function value (For example cos | sin etc.).
I went over everything but my prediction is always straight line
Is it possible and ...
1
vote
1
answer
117
views
catboost ignoring my holiday category
I'm experimenting with catboost, predicting electricity demand from temperature, timeofday, dayofweek and if a day is a public holiday or not (and a few other continuous and categorical columns but ...
- 526
4
votes
1
answer
321
views
Please correct my assumption on how regression trees work
I'm trying to understand how regression trees work, I've been experimenting with catboost and xgboost in python, and I'm getting results which I don't expect, can someone please clarify (and apologies ...
- 526
-1
votes
1
answer
2k
views
What are the key hyperparameters to tune in CatBoost?
I've used XGBoost for a long time but I'm new to CatBoost. If I wanted to run a sklearn RandomizedSearchCV, what are CatBoost's hyperparameters worthwhile including for a binary classification problem?...
- 435
4
votes
1
answer
2k
views
IncToDec Catboost Explained
I am struggling to understand how the overfitting detector with catboost works:
https://tech.yandex.com/catboost/doc/dg/concepts/overfitting-detector-docpage/#overfitting-detector
I am finding ...
- 8,180