3. In the basic decision tree algorithm (assuming we always create binary splits), we choose the fea-ture/value pair with the maximum information gain as the test to use at each internal node of thedecision tree. Suppose we modified the algorithm to choose at random from among those feature/valuecombinations that had non-zero mutual information, and we kept all other parts of the algorithmunchanged.(a) What is the maximum number of leaf nodes that such a decision tree could contain if it weretrained on m training examples?(b) What is the maximum number of leaf nodes that a decision tree could contain if it were trained onm training examples using the original maximum mutual information version of the algorithm?Is it bigger, smaller, or the same as your answer to (b)?(c) How do you think this change (using random splits vs. maximum information mutual informationsplits) would affect the accuracy of the decision trees produced on average? Why?

Decision Tree is a tree that uses a tree like model of decisions and it follows the structure in…

In the basic decision tree algorithm (assuming we always create binary splits), we choose the fea- ture/value pair with the maximum information gain as the test to use at each internal node of the decision tree. Suppose we modified the algorithm to choose at random from among those feature/value combinations that had non-zero mutual information, and we kept all other parts of the algorithm unchanged. (a) What is the maximum number of leaf nodes that such a decision tree could contain if it were

In the basic decision tree algorithm (assuming we always create binary splits), we choose the fea- ture/value pair with the maximum information gain as the test to use at each internal node of the decision tree. Suppose we modified the algorithm to choose at random from among those feature/value combinations that had non-zero mutual information, and we kept all other parts of the algorithm unchanged. (a) What is the maximum number of leaf nodes that such a decision tree could contain if it were

Database System Concepts

7th Edition

ISBN:9780078022159

Author:Abraham Silberschatz Professor, Henry F. Korth, S. Sudarshan

Publisher:Abraham Silberschatz Professor, Henry F. Korth, S. Sudarshan

Chapter1: Introduction

Section: Chapter Questions

Problem 1PE

See similar textbooks

Similar questions

You are given the following game tree. The top level is a maximizing level. A B D E F G 10 12 13 16 13 10 17 6 а. Give the values calculated by minimax for all the nodes labeled by letters in the tree. Do not use alpha-beta pruning. For example, the answer for node D is 12 so you will write D:12 and do the same for all the other nodes. b. Indicate which labeled nodes of the tree will be pruned by alpha-beta pruning.
You perform 2-fold cross-validation to tune the MaxDepth hyperparameter of a decision tree classifier. The MaxDepth hyperparameter can take one of 3 possibles values: {1,2, 3}. These are the accuracies you obtain for each MaxDepth on each fold: Fold MaxDepth 1 3 1 0.9 0.8 0.5 2 0.5 0.8 0.6 What is the MaxDepth hyperparameter chosen by this procedure ? 01 O 2 O 3
In generating a univariate tree, a discrete attribute with n possible val- ues can be represented by n 0/1 dummy variables and then treated as n separate numeric attributes. What are the advantages and disadvan- tages of this approach?
Consider the tree shown above and show the sequence in which the nodes will be visited in an Iterative Deepening Depth-First Search. The algorithm halts when the goal node, node I, is reached. Initially, the depth limit is 0, and node F is visited. Now show all the subsequent iterations.
1 Decision Tree We will use the dataset below to learn a decision tree which predicts the class label which is binary {T,F}. ID A B C 1 1 1 0 0 1 0 F 1 1 1 F 1 1 0 T 2345 Class T 1 F 0 0 • What is the entropy of the root node? Show your working. • Draw the complete decision tree. • Are there feature(s) that are redundant?
Scenario 3: Computing a Tree The algorithm for building phylogenetic trees is O(n^4). Suppose you need to build a tree of 1,000 items. On your computer, it takes 1 minute to build a tree of 200 items. How long will it take to build a tree of the complete data set?
4. The training error of a decision tree with ONE leaf node is 10/40 and 40 is the total number of training samples. What is the pessimistic error of this decision tree: ? After the splitting of the leaf node, there are four leaf nodes, and the training error becomes 9/40. Then what is the pessimistic error of the new decision tree: ? Should you prune the sub-tree after the splitting (yes/no): ?
Consider the 8-puzzle problem. Trace out the A* algorithm given the initial and goal states below. You are supposed to show the generated search tree until the search generates the goal state. For each state show the value of the evaluation function f(n) = g(n) + h(n). Recall that g(n) gives the path cost from the initial state to node n. Moreover, for this problem the step cost is 1 for any action (moving the empty tile). As a heuristic let h(n) be the number of misplaced tiles. Hence f(initial-state) = 0 + 3 = 3. The # symbol denotes the empty tile. Initial State Goal State 1 2 3 1 2 3 # 4 6 4 5 6 7 5 8 7 8 #
Perform Dijkstra’s algorithm (while (Q is not empty){...}, where Q isa min priority queue) on the following graph for source node “s”, writing outevery execution of the RELAX function, the update to the priority queue and theupdates of distance from the source node and parent for the currently visited node.
What is the time complexity for these operation in average case (balanced tree), respectively? What is the time complexity for these operation in average case (balanced tree), respectively? In a BST with n nodes, what is the time complexity for searching, insertion, and deletion in worst case? What is the height of node 3, and node 9, respectively? What is the depth of node 3, and node 9, respectively? What is the level of node 17, and node 11, respectively? Insert integers 11, 22, 15, 44, 6, 9, 3, 33, 17, 5 into a BST in that order. Draw the final tree.
Generate the shortest path tree using the Dijkstra Algorithm using the following format:Shortest path tree for node A1.Permanent list Tentative listempty A(0)2.Permanent list Tentative listA(0) B(5), C(2), D(3)3.Permanent list Tentative listA(0), C(2) B(5), D(3), E(6)4.Permanent list Tentative listA(0), C(2), D(3) B(5), D(3)5.Permanent list Tentative listA(0), C(2), D(3), B(5) D(3)6.Permanent list Tentative listA(0), C(2), D(3), B(5), D(3) empty
Precisely specify the conditions under which the following algorithm returns True, and then discuss, in detail, the running time of the algorithm. If you think it has different best- and worst-case running times then these should be considered separately, and you should explain the conditions under which best and worst-cases arise.