Register

Go Classes | DA Course | Machine Learning

133 views

1 Answer

0 0 votes
Solution: C.

Cross-Entropy loss simplifies to negative $\log$ of the predicted probability for the correct class. At the start of training, we have approximately uniform probabilities for the 10 classes, or 0.1 for each class. So, that means - $\log (0.1)$ is what the loss should approximately be at the start.
 
Voting & Accepting an answer helps improve the community
Answer:
← Previous · Go Classes | DA Cour...Next → · Go Classes | DA Cour...
← Previous in category · Go Classes | DA Cour...Next in category → · Go Classes | DA Cour...
Position:
Show:

Related questions

1 1 vote
1 1 answer
266
266 views
GO Classes asked Mar 18, 2025
266 views
Go Classes | DA Course | Machine Learning
Logistic regression is named after the log-odds of success (the logit of the probability) defined as$$\ln \left(\frac{\mathbb{P}[Y=1 \mid X=x]}{\mathbb{P}[Y=0 \mid X=x]}\...
1 1 vote
0 0 answers
243
243 views
GO Classes asked Mar 18, 2025
243 views
Go Classes | DA Course | Machine Learning
One of the objectives of cross-entropy loss is to both increase the probability of the true class and decrease the probabilities of the other classes. The formula for cro...
1 1 vote
1 1 answer
216
216 views
GO Classes asked Mar 18, 2025
216 views
Go Classes | DA Course | Machine Learning
Consider the following image data point, where the model's predicted probabilities for the classes (Dog, Cat, Mountain) are:\begin{array}{ll}\text { Class } & \text { Pro...
1 1 vote
1 1 answer
217
217 views
GO Classes asked Mar 18, 2025
217 views
Go Classes | DA Course | Machine Learning
Consider the function $f: \mathbb{R}^d \mapsto \mathbb{R}^d$ whose components are given by$$f_i(x)=\frac{x_i^2}{\sum_{k=1}^d x_k^2}$$for $i=1, \ldots, d$ and $x \in \math...