CODEWITHSUNDEEP
catboostcatboostregressor
Avinash Shrivastava
Avinash Shrivastava

Reputation: 81

Usage of class_weights in catboostclassifier

How to use 'class_weights' while using CatboostClassifier for Multiclass problem. The documentation says it should be a list but In what order do I need to put the weights? I have a label array with 15 classes from -2 to +2 including decimal numbers, with class-0 having much higher density compared to the others. Please help. Thanks,

I tried for the binary class which is easier to work with but no clue about multiclass.

cb_model_step1 = run_catboost(X_train, y_train_new, X_test, y_test_new, n_estimators = 1000, verbose=100, eta = 0.3, loss_function = 'MultiClassOneVsAll', class_weights = counter_new)

cb = CatBoostClassifier(thread_count=4, n_estimators=n_estimators, max_depth=10, class_weights = class_weights, eta=eta, loss_function = loss_function)

Upvotes: 8

Views: 18106

Answers (2)

Yaroslav Khnykov
Yaroslav Khnykov

Reputation: 201

Now it is possible to pass a dictionary with labels and corresponding weights.

Suppose we have X_train, y_train and multiclassification problem. Then we can do the following

import numpy as np 
from catboost import CatBoostClassifier
from sklearn.utils.class_weight import compute_class_weight
 
classes = np.unique(y_train)
weights = compute_class_weight(class_weight='balanced', classes=classes, y=y_train)
class_weights = dict(zip(classes, weights))

clf = CatBoostClassifier(loss_function='MultiClassOneVsAll', class_weights=class_weights)
clf.fit(X_train, y_train)

Upvotes: 20

Дмитрий Ушаков
Дмитрий Ушаков

Reputation: 21

you need to fit model without any weights on tour dataset, then run CatBoostClassifier().classes_. it will show you classes order in catboost:

model_multiclass = CatBoostClassifier(iterations=1000,
                       depth=4,
                       learning_rate=0.05,
                       loss_function='MultiClass',
                       verbose=True,
                       early_stopping_rounds = 200,
                       bagging_temperature = 1,
                       metric_period = 100)

model_multiclass.fit(X_train, Y_train)
model_multiclass.classes_
Result:['35мр', '4мр', 'вывод на ИП', 'вывод на кк', 'вывод на фл', 'транзит']

Upvotes: 2

Related Questions