Supervised Learning
Numbers
The assignment is worth 15% of your final grade. Read everything below carefully!
Why?
The purpose of this project is to explore some techniques in supervised learning. It is important to realize that understanding an algorithm or technique requires understanding how it behaves under a variety of circumstances. As such, you will be asked to implement some simple learning algorithms (for sufficiently small values of implement, meaning I dont really want you to implement anything at all), and to compare their performance.
The Means
In this assignment you will go through the process of exploring your chosen datasets, tuning the algorithms you learned about, and writing a thorough analysis of your findings. All that matters is the analysis. It doesnt matter if you implement any learning algorithms yourself (you probably shouldnt) as long as you participate in this journey of exploring, tuning, and analyzing. Concretely, this means you may program in any language that you wish, and you are allowed to use any library you wish as long as it was not written specifically to solve this assignment, and as long as a TA can recreate your experiments on a standard linux machine if necessary (we know how to install a pip package). You might want to look at scikit-learn or Weka, for example.
Some examples of acceptable libraries:
Machine learning algorithms: scikit-learn (python), Weka (java), e107/nnet/random forest(R), ML toolbox (matlab), tensorflow/pytorch (python)
Scientific computing: numpy/scipy(python), Matlab, R
Plotting: matplotlib (python), seaborn (python), Matlab, R
You can use other libraries as long as they fulfill the conditions above. If you are unsure, ask a TA (but please use common sense first)! There is no trick here for you to overthink. Again, the key issue is that I dont care that you implement any of the learning algorithms below; however, I care very much about your analysis.
The Problems Given to You
You should implement five learning algorithms. They are:
Decision trees with some form of pruning
Neural networks
Boosting
Support Vector Machines
k-nearest neighbors
Each algorithm is described in detail in your textbook, the handouts, and all over the web. In fact, instead of implementing the algorithms yourself, you may (and by may I mean should) use software packages that you find elsewhere; however, if you do so you should provide proper attribution. Also, you will note that you have to do some fiddling to get good results, graphs and such, so even if you use anothers package, you may need to be able to modify it in various ways.
Decision Trees. For the decision tree, you should implement or steal a decision tree algorithm (and by implement or steal I mean steal). Be sure to use some form of pruning. You are not required to use information gain (for example, there is something called the GINI index that is sometimes used) to split attributes, but you should describe whatever it is that you do use.
Neural Networks. For the neural network you should implement or steal your favorite kind of network and training algorithm. You may use networks of nodes with as many layers as you like and any activation function you see fit.
Boosting. Implement or steal a boosted version of your decision trees. As before, you will want to use some form of pruning, but presumably because youre using boosting you can afford to be much more aggressive about your pruning.
Support Vector Machines. You should implement (for sufficiently loose definitions of implement including download) SVMs. This should be done in such a way that you can swap out kernel functions. Id like to see at least two.
k-Nearest Neighbors. You should implement (the quotes mean I dont mean it: steal the code) kNN. Use different values of k.
Testing. In addition to implementing (wink), the algorithms described above, you should design two interesting classification problems. For the purposes of this assignment, a classification problem is just a set of training examples and a set of test examples. I dont care where you get the data. You can download some, take some from your own research, or make some up on your own. Be careful about the data you choose, though. Youll have to explain why they are interesting, use them in later assignments, and come to really care about them.
What to Turn In You must submit:
1. a file named README.txt containing instructions for running your code (see note below)
2. a file named yourgtaccount-analysis.pdf containing your writeup
Note below: if the data are way, way, too huge for submitting, see if you can arrange for an URL. This also goes for code, too. Submitting all of Weka isnt necessary, for example, because I can get it myself; however, you should at least submit any files you found necessary to change and enough support and explanation so we could reproduce your results if we really wanted to do so. In any case, include all the information in README.txt
The file yourgtaccount-analysis.pdf should contain:
a description of your classification problems, and why you feel that they are interesting. Think hard about this. To be at all interesting the problems should be non-trivial on the one hand, but capable of admitting comparisons and analysis of the various algorithms on the other.
the training and testing error rates you obtained running the various learning algorithms on your problems. At the very least you should include graphs that show performance on both training and test data as a function of training size (note that this implies that you need to design a classification problem that has more than a trivial amount of data) andfor the algorithms that are iterativetraining times/iterations. Both of these kinds of graphs are referred to as learning curves, BTW.
analyses of your results. Why did you get the results you did? Compare and contrast the different algorithms. What sort of changes might you make to each of those algorithms to improve performance? How fast were they in terms of wall clock time? Iterations? Would cross validation help (and if it would, why didnt you implement it?)? How much performance was due to the problems you chose? How about the values you choose for learning rates, stopping criteria, pruning methods, and so forth (and why doesnt your analysis show results for the different values you chose? Please do look at more than one. And please make sure you understand it, it only counts if the results are meaningful)? Which algorithm performed best? How do you define best? Be creative and think of as many questions you can, and as many answers as you can.
For the sanity of your graders, please keep your analysis as short as possible while still covering the requirements of the assignment: to facilitate this sanity, analysis writeup is limited to 12 pages.
Grading Criteria
You are being graded on your analysis more than anything else. Roughly speaking, implementing everything and getting it to run is worth maybe 0% of the grade on this assignment (I know you dont believe me, but in fact, steal the code; I not only dont
care, I am encouraging you to use one of the many packages available both from the resources page and on the web). Of course, analysis without proof of working code makes the analysis suspect.
The key thing is that your explanations should be both thorough and concise. Imagine you are writing a paper for the major conference in your field the year you will be graduating and you need to impress all those folks who will be deciding whether to interview you later. You dont want them to think youre shallow do you? Or that youre incapable of coming up with interesting classification problems, right? And you surely dont want them to think that you make up for a lack of content by blathering on about irrelevant aspects of your work? Of course not.
Finally, Id like to point out that I am very particular about the format of the assignments. Follow the directions carefully. Failure to turn in files with the proper naming scheme, or anything else that makes the graders lives unduly hard is simply going to lead to an ignored assignment. I am remarkably inflexible about this. Also, there will be no late assignments accepted, so start now. Have fun. One day youll look back on this and smile. There may be tears, but they will be tears of joy.
When your assignment is graded, you will receive feedback explaining your errors (and your successes!) in some level of detail. This feedback is for your benefit, both on this assignment and for future assignments. It is considered a part of your learning goals to internalize this feedback.
If you are convinced that your grade is in error in light of the feedback, you may request a regrade within a week of the grade and feedback being returned to you. A regrade request is only valid if it includes an explanation of where the grader made an error. Send a private Piazza post to only the head TA. In the Summary add [Request] Regrade
It is important to note that because we consider your ability to internalize feedback a learning goal, we also assess it. This ability is considered 10% of each assignment. We default to assigning you full credit. If you request a regrade and do not receive at least 5 points as a result of the request, you will lose those 10 points.
A note about plagiarism and proper citations
Probably the easiest way to fail this class is to plagiarize. Read the note on Piazza when it arrives and make sure it doesnt happen.
BTW, I consider using the code of others in this class to perform the analysis itself to be plagiarism. Way above at the beginning of this assignment I note that I do not care about your implementing machine learning algorithms (and I mean it: I do not believe the learning value is worth it for this course); however, I do care very much that you
understand why your algorithms work and how they are affected by your choice in data and hyper parameters (I believe very, very much in the learning value of this process).
The assignments are actually designed to force you to immerse yourself in the empirical and engineering side of ML that one must master to be a viable practitioner and researcher. The phrase as long as you participate in this journey of exploring, tuning, and analyzing is the key one here. Taking someone elses random seeds, hyper-parameters, and such is, in fact, avoiding the work I care about and circumvents this process because they are the results of the process. Do not circumvent the process.
Reviews
There are no reviews yet.