Student Modelling with an Ensemble Approach


Team Leader

Vladimir Nikulin
University of Queensland
Australia

Overview

Supplementary online material

Provide a URL to a web page, technical memorandum, or a paper.

No response.

Background*

Provide a general summary with relevant background information: Where does the method come from? Is it novel? Name the prior art.

Our method represents a complex of three main directions: random forest in R, Naive Bayes (or averaging), and matrix factorisation, where different KC-opportunities were considered separately. Using above three models we produced 15 different solutions for the test set and for the randomly selected (labelled) subset of the training set, named S, which was set aside and was not used for the primary training. The size of the subset S was about 5% of the whole training data for the algebra and about 2% for the bridge_to_algebra sets. Then, using above 15 solutions as an explanatory variables (or features), we created a secondary training and test sets. The final ensemble solution was constructed by the GBM function in R. We are considering our method as a novel. During last 20 hours of the Challenge we observed a dramatic improvement, and we do not consider our results as a complete or as a final. Our method is a very flexible, and there are a lot of potentials for further improvements and developments.

Method

Summarize the algorithms you used in a way that those skilled in the art should understand what to do. Profile of your methods as follows:

Data exploration and understanding

Did you use data exploration techniques to

  • [not checked]  Identify selection biases
  • [checked]  Identify temporal effects (e.g. students getting better over time)
  • [checked]  Understand the variables
  • [checked]  Explore the usefulness of the KC models
  • [checked]  Understand the relationships between the different KC types

Please describe your data understanding efforts, and interesting observations:

No response.

Preprocessing

Feature generation

  • [not checked]  Features designed to capture the step type (e.g. enter given, or ... )
  • [not checked]  Features based on the textual step name
  • [checked]  Features designed to capture the KC type
  • [checked]  Features based on the textual KC name
  • [checked]  Features derived from opportunity counts
  • [checked]  Features derived from the problem name
  • [checked]  Features based on student ID
  • [not checked]  Other features

Details on feature generation:

No response.

Feature selection

  • [not checked]  Feature ranking with correlation or other criterion (specify below)
  • [not checked]  Filter method (other than feature ranking)
  • [not checked]  Wrapper with forward or backward selection (nested subset method)
  • [not checked]  Wrapper with intensive search (subsets not nested)
  • [not checked]  Embedded method
  • [checked]  Other method not listed above (specify below)

Details on feature selection:

it was rather a feature smoothing based on the method of subintervals

Did you attempt to identify latent factors?

  • [checked]  Cluster students
  • [checked]  Cluster knowledge components
  • [checked]  Cluster steps
  • [checked]  Latent feature discovery was performed jointly with learning

Details on latent factor discovery (techniques used, useful student/step features, how were the factors used, etc.):

also, we used gradient-based matrix factorisation

Other preprocessing

  • [not checked]  Filling missing values (for KC)
  • [not checked]  Principal component analysis

More details on preprocessing:

missings were treated as a special values

Classification

Base classifier

  • [checked]  Decision tree, stub, or Random Forest
  • [checked]  Linear classifier (Fisher's discriminant, SVM, linear regression)
  • [not checked]  Non-linear kernel method (SVM, kernel ridge regression, kernel logistic regression)
  • [checked]  Naïve
  • [not checked]  Bayesian Network (other than Naïve Bayes)
  • [not checked]  Neural Network
  • [not checked]  Bayesian Neural Network
  • [not checked]  Nearest neighbors
  • [checked]  Latent variable models (e.g. matrix factorization)
  • [not checked]  Neighborhood/correlation based collaborative filtering
  • [not checked]  Bayesian Knowledge Tracing
  • [not checked]  Additive Factor Model
  • [not checked]  Item Response Theory
  • [not checked]  Other classifier not listed above (specify below)

Loss Function

  • [not checked]  Hinge loss (like in SVM)
  • [checked]  Square loss (like in ridge regression)
  • [checked]  Logistic loss or cross-entropy (like in logistic regression)
  • [checked]  Exponential loss (like in boosting)
  • [not checked]  None
  • [not checked]  Don't know
  • [not checked]  Other loss (specify below)

Regularizer

  • [not checked]  One-norm (sum of weight magnitudes, like in Lasso)
  • [not checked]  Two-norm (||w||^2, like in ridge regression and regular SVM)
  • [not checked]  Structured regularizer (like in group lasso)
  • [not checked]  None
  • [not checked]  Don't know
  • [checked]  Other (specify below)

Ensemble Method

  • [checked]  Boosting
  • [not checked]  Bagging (check this if you use Random Forest)
  • [checked]  Other ensemble method
  • [not checked]  None

Were you able to use information present only in the training set?

  • [checked]  Corrects, incorrects, hints
  • [not checked]  Step start/end times

Did you use post-training calibration to obtain accurate probabilities?

  • [selected]  Yes
  • [not selected]  No

Did you make use of the development data sets for training?

  • [selected]  Yes
  • [not selected]  No

Details on classification:

used k-means clustering applied to the latent factors with regularisation. The target of the regularisation is to ensure that all the clusters will be sufficiently large.

Model selection/hyperparameter selection

  • [checked]  We used the online feedback of the leaderboard.
  • [checked]  K-fold or leave-one-out cross-validation (using training data)
  • [not checked]  Virtual leave-one-out (closed for estimations of LOO with a single classifier training)
  • [not checked]  Out-of-bag estimation (for bagging methods)
  • [not checked]  Bootstrap estimation (other than out-of-bag)
  • [not checked]  Other cross-validation method
  • [not checked]  Bayesian model selection
  • [not checked]  Penalty-based method (non-Bayesian)
  • [not checked]  Bi-level optimization
  • [not checked]  Other method not listed above (specify below)

Details on model selection:

No response.

Results

Final Team Submission

Scores shown in the table below are Cup scores, not leaderboard scores. The difference between the two is described on the Evaluation page.

A reader should also know from reading the fact sheet what the strength of the method is.

Please comment about the following:

Quantitative advantages (e.g., compact feature subset, simplicity, computational advantages).

simplicity of the model is a very important when data are large and complex.

Qualitative advantages (e.g. compute posterior probabilities, theoretically motivated, has some elements of novelty).

We do believe that our method has quite interesting and novel theoretical grounds.

Other methods. List other methods you tried.

No response.

How helpful did you find the included KC models?

  • [selected]  Crucial in getting good predictions
  • [not selected]  Somewhat helpful in getting good predictions
  • [not selected]  Neutral
  • [not selected]  Not particularly helpful
  • [not selected]  Irrelevant

If you learned latent factors, how helpful were they?

  • [selected]  Crucial in getting good predictions
  • [not selected]  Somewhat helpful in getting good predictions
  • [not selected]  Neutral
  • [not selected]  Not particularly helpful
  • [not selected]  Irrelevant

Details on the relevance of the KC models and latent factors:

No response.

Software Implementation

Availability

  • [checked]  Proprietary in-house software
  • [not checked]  Commercially available in-house software
  • [not checked]  Freeware or shareware in-house software
  • [not checked]  Off-the-shelf third party commercial software
  • [checked]  Off-the-shelf third party freeware or shareware

Language

  • [checked]  C/C++
  • [not checked]  Java
  • [checked]  Matlab
  • [not checked]  Python/NumPy/SciPy
  • [checked]  Other (specify below)

Details on software implementation:

R and Perl

Hardware implementation

Platform

  • [not checked]  Windows
  • [checked]  Linux or other Unix
  • [not checked]  Mac OS
  • [not checked]  Other (specify below)

Memory

  • [not selected]  <= 2 GB
  • [not selected]  <= 8 GB
  • [selected]  >= 8 GB
  • [not selected]  >= 32 GB

Parallelism

  • [checked]  Multi-processor machine
  • [not checked]  Run in parallel different algorithms on different machines
  • [not checked]  Other (specify below)

Details on hardware implementation. Specify whether you provide a self contained-application or libraries.

No response.

Code URL

Provide a URL for the code (if available):

No response.

Competition Setup

From a performance point of view, the training set was

  • [selected]  Too big (could have achieved the same performance with significantly less data)
  • [not selected]  Too small (more data would have led to better performance)

From a computational point of view, the training set was

  • [not selected]  Too big (imposed serious computational challenges, limited the types of methods that can be applied)
  • [selected]  Adequate (the computational load was easy to handle)

Was the time constraint imposed by the challenge a difficulty or did you feel enough time to understand the data, prepare it, and train models?

  • [selected]  Not enough time
  • [not selected]  Enough time
  • [not selected]  It was enough time to do something decent, but there was a lot left to explore. With more time performance could have been significantly improved.

How likely are you to keep working on this problem?

  • [not selected]  It is my main research area.
  • [selected]  It was a very interesting problem. I'll keep working on it.
  • [not selected]  This data is a good fit for the data mining methods I am using/developing. I will use it in the future for empirical evaluation.
  • [not selected]  Maybe I'll try some ideas , but it is not high priority.
  • [not selected]  Not likely to keep working on it.

Comments on the problem (What aspects of the problem you found most interesting? Did it inspire you to develop new techniques?)

Yes, the problem is very interesting, and inspired us to develop a lot of new software (written in C). However, we had a lot of problems with pre-processings, which were conducted using Perl, and we can not be sure that all the problems have been resolved. We even requested the Organisers of the Cup (about one week before the deadline) to make an extension for one week.

References

List references below.

No response.