1. Explain the curse of dimensionality. The report has to answer (at least) what it means from the perspective of machine learning, what problems it causes, and how it can be resolved or mitigated.

2. Explain what regularization and Occam's razor are and how they are related. Answer two regularization methods for linear regression and discuss their pros and cons with the reasons.

3. Explain the method of Lagrange multipliers. The report has to answer (at least) the class of problems it can solve and the algorithm (i.e., how it solves the problems).

4. Explain with mathematical description why LSTMs can mitigate the vanishing gradient problem.

5. Explain AlphaGo, one of the state-of-the-art machine learning models that apply neural networks to reinforcement learning tasks. The report has to answer (at least) what problem it solves, what its architecture is, and how it works.