assignments:assignment2
Differences
This shows you the differences between two versions of the page.
| Both sides previous revision Previous revision Next revision | Previous revision Last revision Both sides next revision | ||
|
assignments:assignment2 [2016/09/06 09:51] asa |
assignments:assignment2 [2016/09/07 13:22] asa |
||
|---|---|---|---|
| Line 61: | Line 61: | ||
| The variable $\eta$ plays the role of the learning rate $\eta$ employed in the perceptron algorithm and $\delta \alpha$ is the proposed magnitude of change in $\alpha_i$. | The variable $\eta$ plays the role of the learning rate $\eta$ employed in the perceptron algorithm and $\delta \alpha$ is the proposed magnitude of change in $\alpha_i$. | ||
| We note that the adatron tries to maintain a //sparse// representation in terms of the training examples by keeping many $\alpha_i$ equal to zero. The adatron converges to a special case of the SVM algorithm that we will learn later in the semester; this algorithm tries to maximize the margin with which each example is classified, which is captured by the variable $\gamma$ in the algorithm (notice that the magnitude of change proposed for each $\alpha_i$ becomes smaller as the margin increases towards 1). | We note that the adatron tries to maintain a //sparse// representation in terms of the training examples by keeping many $\alpha_i$ equal to zero. The adatron converges to a special case of the SVM algorithm that we will learn later in the semester; this algorithm tries to maximize the margin with which each example is classified, which is captured by the variable $\gamma$ in the algorithm (notice that the magnitude of change proposed for each $\alpha_i$ becomes smaller as the margin increases towards 1). | ||
| + | |||
| + | **Note:** if you observe an overflow issues in running the adatron, add an upper bound on the value of $\alpha_i$. | ||
| Here's what you need to do: | Here's what you need to do: | ||
| - | - Implement the pocket algorithm and the adatron; each classifier should be implemented by a separate class, and use the same interface used in the code provided for the perceptron algorithm. Make sure each classifier you use (including the original version of the perceptron) implements a bias term. | + | - Implement the pocket algorithm and the adatron; each classifier should be implemented in a separate Python class (they can all be in the same module), and use the same interface used in the code provided for the perceptron algorithm, i.e. provides the same methods. Make sure each classifier you use (including the original version of the perceptron) implements a bias term. |
| - | - Compare the performance of these variants of the perceptron on the Gisette and QSAR datasets by computing an estimate of the out of sample error on a sample of the data that you reserve for testing (the test set). In each case reserve about 60% of the data for training, and 40% for testing. To gain more confidence in our error estimates, repeat this experiment using 10 random splits of the data into training/test sets. Report the average error and its standard deviation in a [[https://en.wikibooks.org/wiki/LaTeX/Tables|LaTex table]]. Is there a version of the perceptron that appears to perform better? (In answering this, consider the differences you observe in comparison to the standard deviation). | + | - Compare the performance of these variants of the perceptron on the Gisette and QSAR datasets by computing an estimate of the out of sample error on a sample of the data that you reserve for testing (the test set). In each case reserve about 60% of the data for training, and 40% for testing. To gain more confidence in our error estimates, repeat this experiment using 10 random splits of the data into training/test sets. Report the average error and its standard deviation in a [[https://en.wikibooks.org/wiki/LaTeX/Tables|LaTex table]]. Is there a version of the perceptron that appears to perform better? (In answering this, consider the differences in performance you observe in comparison to the standard deviation). |
assignments/assignment2.txt · Last modified: 2016/09/14 09:38 by asa
Page Tools
Except where otherwise noted, content on this wiki is licensed under the following license: CC Attribution-Share Alike 4.0 International
