ML4T Exam 1 ACTUAL UPDATED QUESTIONS AND CORRECT ANSWERS
MDP Markov Decision Process: inifinte horizon means lasts forever.
EMH Efficient Market Hypothesis asserts that financial markets are informationally
efficient, and so one cannot consistently earn returns in excess of average market
returns on a risk adjusted basis ("no arbitrage" assumption). In the EMH, price is
the best measurement of value.
Synthetic CDO bet on real CDOS. its made of bets of credit default swaps
Dyna reinforcement learning framework.
Learn from real experience (like Q-learning).
Update a model of the environment (learns transitions & rewards).
Use the model to simulate fake experiences (planning), and improve the policy
faster.
Q Learning Q-learning is a type of Reinforcement Learning algorithm where an agent learns
how to act by trying things out and getting feedback in the form of rewards.
It learns what actions are best to take in each situation without needing to know
how the environment works in advance.
Q stands for "quality" of an action in a given state.
The algorithm learns a table called the Q-table, where each entry Q(s,a)Q(s,
a)Q(s,a) represents the expected future reward from taking action a in state s, and
then acting optimally after that.
MDP Markov Decision Process: inifinte horizon means lasts forever.
EMH Efficient Market Hypothesis asserts that financial markets are informationally
efficient, and so one cannot consistently earn returns in excess of average market
returns on a risk adjusted basis ("no arbitrage" assumption). In the EMH, price is
the best measurement of value.
Synthetic CDO bet on real CDOS. its made of bets of credit default swaps
Dyna reinforcement learning framework.
Learn from real experience (like Q-learning).
Update a model of the environment (learns transitions & rewards).
Use the model to simulate fake experiences (planning), and improve the policy
faster.
Q Learning Q-learning is a type of Reinforcement Learning algorithm where an agent learns
how to act by trying things out and getting feedback in the form of rewards.
It learns what actions are best to take in each situation without needing to know
how the environment works in advance.
Q stands for "quality" of an action in a given state.
The algorithm learns a table called the Q-table, where each entry Q(s,a)Q(s,
a)Q(s,a) represents the expected future reward from taking action a in state s, and
then acting optimally after that.
