Introduction from the instructor
What is feature selection?
Many materials are from an earlier course by the competition organizers.
Finish reading the report and repeat their baseline and best entries. You need to install CLOP. The easiest way is to check their quick start guide.
Present your experiences. The presentation doesn't need to be formal. You can highlight your experimental log, and may show some additional stuff you have tried. Moreover, it is good to raise questions here for discussion. You may also want to mention difficulties in using CLOP, or even have suggestions to the authors.
Details of the five preprocessors, five feature selection methods, and five other feature extraction classes are presented.
Discuss how students come up with the best entry
How to achieve the best performance without trying too many submissions? Show your experimental results.
Prepare results for the first and the second data sets. We will submit them to the organizers.
Before the submission, each of you needs to present your experiments and explain your final approach.
Prepare results for the third to the fifth data sets. We will submit them to the organizers.
Before the submission, each of you needs to present your experiments and explain your final approach.