CS 66 Lab 8

Due Monday, 11/14/2022, by midnight (23:59, EST)

Announcements

Goals

The goals for this lab assignment are:

  • Learn how to compare machine learning algorithms

  • Practice your final paper’s programming part.

  • Reproduce machine learning results from classic papers

1. Reviewer Guide

  • Pay attention to the section of 'Review content;

  • Considering the reviewer guide when preparing your labs and final paper

    1. Summary and contributions:

    2. Strengths:

    3. Weaknesses:

    4. Correctness:

    5. Clarity:

    6. Relation to prior work:

    7. Reproducibility:

    8. Additional feedback:

    9. Overall score:

    10. Confidence score:

    11. Broader impact:

    12. Ethical concerns:

2. Paper Example

Please read this example below:

  • EEGEyeNet

  • Reproduce the left-right task, and track the runtime for each algorithm.

  • Write a result section to summarize your findings, with a result table.

3. CNN Examples

Please read this example page below:

  • MNIST dataset

  • Reproduce results of at least two single classifiers, such as KNN and SVM

  • Investigate and reproduce results of at least two ensemble methods, such as Random Forest and boosting

  • Reproduce results of CNN and at least one of its variations.

  • (Optional) Investigate and reproduce results of two newer algorims, such xgboost and lightGBM.

  • Track the runtime for each algorithm you reproduce.

  • Write a result section to summarize your findings, with a result table.

4. Submission Guide

  • This is an team assignment.

  • Each team submits one file, lab_8_lastname.zip, including

    1. lab_8_code.zip, file size maximum is 3M.

    2. lab_8_Results.pdf. Your results for these two experiments.

5. Notes

  • Email 'xqu1@swarthmore.edu' your zip file for lab 8.

  • please get in touch with the instructor and the teaching assistant ASAP if you encounter programming difficulties.