search

CONECT-INT / 2023_CENTURI-SummerSchool

Repository for the 2023 CENTURI Summer School project on neural decoding
GNU General Public License v3.0
2 stars 1 forks source link

testing sigmoid vs soft max in dataset1 #8

Open laurentperrinet opened 1 year ago

laurentperrinet commented 1 year ago

from https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html#sklearn.linear_model.LogisticRegression

In the multiclass case, the training algorithm uses the one-vs-rest (OvR) scheme if the ‘multi_class’ option is set to ‘ovr’, and uses the cross-entropy loss if the ‘multi_class’ option is set to ‘multinomial’. (Currently the ‘multinomial’ option is supported only by the ‘lbfgs’, ‘sag’, ‘saga’ and ‘newton-cg’ solvers.)

it would be useful to test ‘ multinomial’ (sigmoid) vs ‘ovr’ (softmax)

laurentperrinet commented 1 year ago

checkout https://sebastianraschka.com/blog/2022/losses-learned-part1.html for a nice introduction...