CA-3 KM062-G1-Even

INT-234

Predictive Analytics

Roll No:____________ Section:________________

Note:

1. All questions are compulsory.

2. Copy cases will be awarded zero without any explanation.

3. Each question is of 15 marks.

4. Write all answers on paper and also paste the screenshot of code and output of every
statement.

Q1. Use the Wholesale Customers data from Berkley UCI Machine Learning Repository and
divide the data into 5 different clusters using K Means Clustering.

Q2. Use the cererals.csv file. Find the rating of the cereals depending upon the given parameters.
Train your machine with 60% data using Neural Network. Use 3 hidden nodes in your network.
Also plot the neural network and find the error.
 CA-3 KM062-G1-Odd

INT-234

Predictive Analytics

Roll No:____________ Section:________________

Note:

1. All questions are compulsory.

2. Copy cases will be awarded zero without any explanation.

3. Each question is of 15 marks.

4. Write all answers on paper and also paste the screenshot of code and output of every
statement.

Q1. Use Heart disease dataset and implement SVM on the same. Train your machine with 80%
data so that it can predict whether a person is suffering from any heart disease or not. Also create
the confusion matrix.

Q2. Use the College dataset available in ISLR package. Train machine with 70% data using
neural network so that it can predict whether a college is private or not using all the given
parameters. Use hidden = c(10,10,10) and plot the neural network. Create a confusion matrix
also.
 CA-3 KM062-G2-Odd

INT-234

Predictive Analytics

Roll No:____________ Section:________________

Note:

1. All questions are compulsory.

2. Copy cases will be awarded zero without any explanation.

3. Each question is of 15 marks.

4. Write all answers on paper and also paste the screenshot of code and output of every
statement.

Q1. Use the glass data from the UCI Repository of Machine Learning Databases. Predict the type
of glass on basis of its chemical analysis using SVM.

Q2. Use the carseats dataset available in ISLR package. Train machine with 70% data using
neural network so that it can predict sales depending on income, advertising, population, price
and age. Use hidden = c(7,6,5) and plot the neural network. Create a confusion matrix also.
 CA-3 KM062-G2-Even

INT-234

Predictive Analytics

Roll No:____________ Section:________________

Note:

1. All questions are compulsory.

2. Copy cases will be awarded zero without any explanation.

3. Each question is of 15 marks.

4. Write all answers on paper and also paste the screenshot of code and output of every
statement.

Q1. Use built-in USArrests dataset which contains arrests per 100000 residents for assault,
murder and rape in each of the 50 US states in 1973. Use K means clustering to divide the whole
data into 2 clusters.

Q2. Download dividendinfo.csv from github. Train machine with 80% data using Neural
network and predict the dividend depending on other given parameters. Plot the neural network
and also create the confusion matrix.