The dataset - called Labeled Faces in the Wild - contains pictures of celebrities. You will be using a subset containing five celebrities, each of whom has at least 80 pictures. Requirements: 1. Build prediction models using LogisticRegression and SVC. You must try out different things to get the best estimator. suggest that you use GridSearchCV or RandomSearchCV to find the best estimator.

This is an imbalanced dataset, but you are not required to use oversampling or undersampling.

2. Build a prediction model using the deep-learning concepts that have been covered so far. Experiment with layers, hidden units within the layers, batch size, optimizers, and learning rate, and document your results.

Compare the performance of your deep learning model with that of the models you used in 1.

I need an actual code, not just the answer. This is the starting Python code. Thanks in advance.