firm 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38wc 3338 W r i t i n g

This assessment item relates to the unit learning outcomes as in the unit descriptor. This assessment is designed to improve student collaborative skills in a team environment and to give students experience in constructing a range of documents as deliverables form different stages of the Intelligent Systems for Analytics

INSTRUCTIONS

Assignment 3 :- Group Assignment (30 %) and submission at week 12

In this assignment students will work in small groups to develop components of the Documents discussed in lectures. Student groups should be formed by Session four. Each group needs to complete the group participation form attached to the end of this document. Assignments will not be graded unless the student has signed a group participation form.

Carefully read the following two questions and provide the appropriate answer.

Question 1. The bankruptcy-prediction problem can be viewed as a problem of classification. The data set you will be using for this problem includes two ratios that have been computed from the financial statements of real-world firms. These two ratios have been used in studies involving bankruptcy prediction. The first sample (training set) includes 68 data value on firms that went bankrupt and firms that didn’t. This will be your training sample. The second sample (testing set) of 68 firms also consists of some bankrupt firms and some non bankrupt firms. Your goal is to use different classifiers to build a training model, by randomly selecting the 40 data points (20 points from category 1 and 20 points from category 0), and then test its performance on the testing model by randomly selecting 40 data points from the testing set. (Try to analyze the new cases yourself manually before you run the neural network and see how well you do). Both Data Sets are provided below:

Students have to use the following classifiers. The selection of the classifiers depend upon the members of the group. E.g. If the group has four members then they will use the four classifiers from the following six classifiers.

1. Neural networks

2. Support vector machines

3. Nearest neighbor algorithms

4. Decision trees

5. Naive Bayes

6. Any other classifier

MITS5509 Assignment 3

The following tables show the training sample and test data you should use for this exercise.

Firm 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38

WC 3338.61

3801.72 2818.817 1250.953 2444.406 937.917 1600.792 3128.813 2486.803 4220.996 2585.41 3512.085 4170.333 938.879 1437.695 627.985 4430.049 989.568 3275.474 1500.437 848.989 1386.494 1554.257 2228.338 2568.391 1720.128 4106.106 3500.883 1217.846 3544.406 2082.873 709.01 2523.939 2781.307 309.577 363.79 341.399 363.616

Training set

DC 0.56555 0.570567 0.572058 0.568258 0.553276 0.561066 0.534662 0.564714 0.564239 0.58465 0.572457 0.550878 0.569516 0.545574 0.529922 0.51941 0.567547 0.534501 0.555306 0.565886 0.548603 0.56229 0.562346 0.565556 0.54973 0.568458 0.57767 0.557197 0.525333 0.568735 0.557527 0.541673 0.55366 0.569188 0.557668 0.561751 0.550717 0.568882

Category 1