Sale!

Machine Learning- How do Ensembles Work?

$30.00

Category:

Machine Learning
Worth: 5% of total grade [5 marks]

How do Ensembles Work? (5 marks):
You will have to run the Experimenter in Weka. You will need to run Randomforest,
AdaBoostM1, and Bagging. You will have to run all three algorithms on 3 datasets: Iris,
Car, and Balance-scale. Make sure you run 10-fold cross validation so you can tell which
algorithm is better on which dataset. Run the default settings on the 3 algorithms. It is
important that you input the datasets in this order:
1. iris.arff
2. car.arff
3. balance-scale.arff
It is important you enter the algorithms in this order:
1. RandomForest
2. AdaBoostM1
3. Bagging
This is to make the marking of your assignments more automated. If you do them in
another order you might not get full marks. In addition it is important to use the default
settings for all the algorithms, please do not change them.
What you need to turn in:
You need to turn in a .pdf file of 1 page MAXIMUM which should include:
1) The two-tailed t-test (corrected) output from Weka “Analyse” in the
Experimenter. Include a screen shot of this in your .pdf
2) A paragraph about each of the 3 datasets saying why you think a particular
algorithm did the best of each dataset.
3) A final paragraph saying which algorithm you think is usually better and why.
Marking is based on the following material:
½ mark for having the correct Weka output
3 marks (1 for each description of the 3 datasets)
1 mark for the final paragraph saying which algorithm is more reliable
½ mark for good presentation and English
The datasets can be found at
https://canvas.auckland.ac.nz/courses/40682/files/folder/Assignment2
The assignment must be submitted to Canvas. It will be run through Turnitin.

Reviews

There are no reviews yet.

Be the first to review “Machine Learning- How do Ensembles Work?”

Your email address will not be published.