1. (100 points) The following neural network is used for mortgage approval. Char

Question

1. (100 points) The following neural network is used for mortgage approval. Character .5 Mortgage Fina Age greep Credit A The numbers 1-7 identify nodes, and the weights take positive or negative integer values. For example, the weight for the connection between Node 1 and Node 5 is -5. Linear activation functions are used for the input nodes, and binary threshold activation functions, which are set to zero, for the hidden and output nodes. The components of an input pattern correspond to four attributes, namely [character, mortgage, age.group, credit.conditionl and takes values in the following sets: character (good bad mortgage-usano ; age.croun= { adult.cniddle-aged); sredit.condition (good, bad)

Explanation / Answer

Hi,

Since we are given the character is good so we will send {1,0} to node character which is node 1.

Also, we are given the no for mortgage so we will send {0,1} to node mortgage which is node 2.

Similarly, we will send {1,0} to node age_group which is node 3 and we will send {0,1} to node credit_condition which is node 4

So, on node 5 we will recieve -5*1 + 3*0 + 2*1 + 4*0 = -5 + 0 + 2 + 0 = -3

This is computed by the fact that output in form of 1 and 0 is being generated by first node, {1,0} in case of good character and {0,1} in case of bad character. The generated pair is sent in the proceeding branches and the branch corresponsing to 0 is taken as false and 1 is taken as true since this is a conditional statement.

So, in case of good character when node 1 generates {1,0} ,So, 1 multiplied by (-5) which is weight of branch is sent forward i.e. to node 5 i.e. -5 is sent to node 5 in case of good character and 0 multiplied by 6 which is weight of another branch is sent forward i.e. to node 6 i.e. 0 is sent to node 6 in case of bad character .

In the similar way, we can calculate what is being sent to node 5 and node 6 by node 2, node 3 and node 4 provided the given input.

So, we have computed the output of node 5 is -3.

In the similar way, let's calculate the output of node 6

6*0 + (-1)*1 + (-2)*0 + 5*1 = 0 + (-1) + 0 + 5 = 4

So, the output of node 6 is 4.

Now, as we see from neural nerwork diagram, the branch pointing from node 5 to node 7 has -1 weight and branch providing from 6 to 7 has 2 weight. So, output of node 7 is :

output of node5 * wieght of branch from node5 to node7 + output of node6 * wieght of branch from node6 to node7 = (-3)*(-1) + 4*2 = 3 + 8 =11. So, output is 11.

Hope this will help you!!! Thanks!!! Do give this a thumbs up!!!:)

Related Questions

Navigate

• Browse (All)
• Browse 1
• Subjects

---

---

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