When two classes of data are linearly separable, and infinite number of hyperpla

Question

When two classes of data are linearly separable, and infinite number of hyperplanes (decision boundaries) can be learned from that data. The 'best' hyper plane is chosen as the one that Minimizes classification error on the training data Minimizes classification error on the test data Has the least value for its gradient Leaves the maximum margin from training data belonging to the two classes Breadth-first search Expands the shallowest nodes first Expands the node lowest path cost Expands the deepest unexpanded node first None of the above Uniform-cost search Expands the shallowest nodes first Expands the node lowest path cost Expands the deepest unexpanded node first None of the above Depth-first search Expands the shallowest nodes first Expands the node lowest path cost Expands the deepest unexpanded node first None of the above Iterative deepening search calls uniform-cost search with increasing depth limits until a goal is found (True/False) A heuristic function h(n) estimates the cost already paid to reach the node n estimates the cost of a solution as a function of a number of iterations n estimates the cost of a solution from node n estimates the cost of a solution from the start node to the goal node n In genetic algorithms new states are generated by Mutation only Mutation and crossover Crossover only A hill-climbing algorithm

Explanation / Answer

14.

d) Leaves the maximum margin from training data belonging to the two classes.

15. b) Expands the ndoe with lowest path cost

16. a) Expands the shallowest nodes first

17. c) Expands the deepest unexpected ndoe first

Related Questions

Navigate

• Browse (All)
• Browse W
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.