Blog

Suppose there is a blocksworld domain that contains some blocks and a table. A block can be on top of the table or on the other block. On relation specifies which block is on top of what. Move action moves a block from one location to another. In_Gripper relationship specifies that the block is in the gripper. Consider these states: Current State: block(b1), block(b2), block(b3), block(b4), On(b1,b2), On(b2,table), On(b3,table), On(b4,table) Goal State: On(b2,table), On(b1,b2), On(b3,b4) In order for a state to be a landmark, which proposition must be contained in the state?

What capability does structure from motion (SfM) contribute to robotic perception?

Table: Gridworld MDP Table: Gridworld MDP   Figure: Transition Function Figure: Transition Function   Review Table: Gridworld MDP and Figure: Transition Function. The gridworld MDP operates like the one discussed in lecture. The states are grid squares, identified by their column (A, B, or C) and row (1 or 2) values, as presented in the table. The agent always starts in state (A,1), marked with the letter S. There are two terminal goal states: (B,1) with reward -5, and (B,2) with reward +5. Rewards are 0 in non-terminal states. (The reward for a state is received before the agent applies the next action.) The transition function in Figure: Transition Function is such that the intended agent movement (Up, Down, Left, or Right) happens with probability 0.8. The probability that the agent ends up in one of the states perpendicular to the intended direction is 0.1 each. If a collision with a wall happens, the agent stays in the same state, and the drift probability is added to the probability of remaining in the same state. Assume that V1_1(A,1) = 0, V1_1​(C,1) = 0, V1_1​(C,2) = 4, V1_1​(A,2) = 4, V1_1​(B,1) = -5, and V1_1​(B,2) = +5. Given this information, what is the second round of value iteration (V2_2) update for state (A,1) with a discount of 1?

How does structure from motion (SfM) predict 3D structures?

Consider a logistic domain where there are 5 cities, 50 trucks, and 50 packages. Each truck can be at any of the cities. A package can either be at one of the cities or in one of the trucks. Assuming that a truck can go from any city to any other city, what is the minimum number of variables that will be needed to represent this problem, using factored representation?

Consider a logistic domain where there are 100 cities, 5 trucks, and 6 packages. Each truck can be at any of the cities. A package can either be at one of the cities or in one of the trucks. Assuming that a truck can go from any city to any other city, what is the minimum number of variables that will be needed to represent this problem, using factored representation?

Consider a logistic domain where there are 2 cities, 2 trucks, 3 drivers, and 2 packages. Each truck or driver or package can be in any of the cities. If atomic representation is used, what would be the minimum number of states?

Table: Gridworld MDP Table: Gridworld MDP   Figure: Transition Function Figure: Transition Function   Review Table: Gridworld MDP and Figure: Transition Function. The gridworld MDP operates like the one discussed in lecture. The states are grid squares, identified by their column (A, B, or C) and row (1 or 2) values, as presented in the table. The agent always starts in state (A,1), marked with the letter S. There are two terminal goal states: (C,2) with reward +1, and (A,2) with reward -1. Rewards are 0 in non-terminal states. (The reward for a state is received before the agent applies the next action.) The transition function in Figure: Transition Function is such that the intended agent movement (Up, Down, Left, or Right) happens with probability 0.8. The probability that the agent ends up in one of the states perpendicular to the intended direction is 0.1 each. If a collision with a wall happens, the agent stays in the same state, and the drift probability is added to the probability of remaining in the same state. The discounting factor is 1. The agent starts with the policy that always chooses to go Up, and it executes three trials: the first trial is (A,1)–(A,2), the second is (A,1)–(A,2), and the third is (A,1)–(B,1)–(C,1)–(C,2). Given these traces, what is the Monte Carlo (direct utility) estimate for state (A,1)?

Actions: (:action moveTruck :parameters(?t – truck ?source_loc ?dest_loc – location) :precondition(and (truck_at ?t ?source_loc) (path ?source_loc ?dest_loc)) :effect(and (not (truck_at ?t ?source_loc)(truck_at ?loc ?dest_loc)) ) (:action load :parameters(?p – package ?t – truck ?loc – location) :precondition(and (package_at ?p ?loc)(truck_at ?t ?loc)) :effect(and (not (packeg_at ?p ?loc))(in ?p ?t)) ) (:action unload :parameters(?p – package ?t – truck ?loc – location) :precondition(and (truck_at ?t ?loc)(in ?p ?t)) :effect(and (not (in ?p ?t))(package_at ?p ?loc)) ) Current State: (truck_at truck_2 location_1)(truck_at truck_1 location_2)(package_at package_1 location_1)(package_at package_2 location_2)(path location_1 location_2)(path location_2 location_1) Consider the provided action descriptions and current state. Given this information, which action can be executed?

Suppose you are training a neural network. One of the neurons has an input vector [12, 35, -80, -0.4, 21] , weights [0.1, -0.6, 0.2, -0.01, 2] , bias 0.45, and a ReLu activation function. Which value is the output of the ReLu activation function?