I am repeatedly surprised by how often these three things appear while any ML discussion is there:
- Log-Likelihood: I understand the max likelihood principle, why log?
- Softmax: Why softmax everywhere? Is it tied to log-likelihood in any way.
- Sigmoid: Why sigmoid function only in NNs?
Please help me understand/direct me to resources which provided an intuitive + mathematical (rigorous) validations of these observations. Thanks a lot.