PBL (Project Based Learning)

Final project

Project Call

From Problem Definition to Insightful

In this project, students will embark on a hands-on journey through the complete machine learning pipeline. The goal is to apply the knowledge gained in class to define a real-world problem, collect relevant data, preprocess it, and perform data analysis either to (1) predict a target variable (supervised learning) or (2) uncover patterns in the data (unsupervised learning). The emphasis is on understanding the logical process from problem identification to generating insights and effectively telling the story of your analysis in a final report.

• Objectives: (related to score)

  • To develop a clear and well-defined problem statement.

  • To understand the data collection process and gather data relevant to the problem.

  • To apply appropriate data preprocessing techniques to clean and prepare the data for analysis.

  • To explore the data using machine learning techniques for prediction or pattern discovery.

  • To communicate findings and the logical reasoning behind them in a structured, compelling report.

• Deliverable: Submit a final report that includes all steps of your project. Any format is possible (PPT, word, notion web page link, pdf, and so on). Your report should include:

  • Introduction: Problem definition and context.

  • Data Description: Dataset, source, and preprocessing steps.

  • Analysis: Methods, models, or algorithms used, and results.

  • Conclusion: Insights, limitations, and potential next steps.

  • Appendix: Any additional resources, code snippets, or references.

  • Final Presentation

    • YouTube Video Submission: Submit a 10-minute YouTube video (in any language). The video link should be shared along with the final report.

    • Due date: 16 Dec (Tue) 09:00 (Before the class)

    • The best teams may lead to URP in the summer semester.

• Grading criteria

  • Problem Definition (10%): Clarity, relevance, and impact of the problem statement.

  • Data Collection (10%): Appropriateness and quality of the data, description of the data collection process.

  • Data Preprocessing (15%): Completeness of data cleaning, appropriateness of techniques, and explanation of decisions.

  • Data Analysis (30%): Correctness of the methods used, depth of analysis, and justification of conclusions.

  • Final Report (20%): Quality of storytelling, structure, and clarity of presentation, visualizations, and overall communication.

  • Presentation (15%): Quality of communication to audiances

• Students organize teams that meet several conditions.

  • 4~5 members in a team

  • Background diversity: no homogeneous majors in a team

  • Exception: Allowed if persuasion is possible for sufficient reasons

  • Use Kakao openchat room to recruit your team members

• Teams

  A	B	C	D	E	G	L	M