Write True if the statement is True; False, otherwise.

2.1. A* will generally be more efficient than Uniform Cost search since, in the worst case, an admissible heuristic gives no additional help but it never hurts.

2.2. It is impossible to over-train a multi-layer feed-forward network using the back-propagation learning algorithm. It is guaranteed that the longer you train your system, the more accurate it will perform.

2.3. The use of the Visited List improves the performance of Uniform Cost search without affecting its correctness.

2.4. If constraint propagation leaves some variable with an empty domain, there is no solution.

2.5. Iterative deepening combined with alpha-beta pruning is suitable for building a computer program that plays chess because the results of previous searches can be used to order the moves. Previous search results can also be used for extracting better initial values for alpha and beta for the next successive iterations of the search. This will help us cut-off irrelevant moves early.

2.6. A training set is used for estimating how a Neural Network performs in the real-world.

2.7. Softmax units are used especially for solving regression problems, provided that the sum of the output units is equal to 1.

2.8. The Alpha-beta algorithm will return exactly the same result as the Min-Max algorithm.

Attachment:- assignment.pdf