Student Presentations for Exploratory Data Analysis (EDA): House Price Forecast

Machine Learning for Engineers using Orange3 without coding

Student Presentations for Exploratory Data Analysis (EDA): Heart Stroke Forecast

COMPE468 Machine Learning For Engineers
Murat Karakaya

Exploratory Data Analysis (EDA) & Feature Engineering Guideline

You can check the following lecture note on Moodle and the blog posts:

Consider checking the previous years’ presentations, reports, and recordings on Moodle so you do not miss important actions.
IMPORTANT NOTES:
• You DO NOT need to code anything! You can implement a similar analysis in the above posts using Orang3 widgets- they are exactly from the same Python library.
• However, please note that creating tens of figures is not our aim!!!
• Our goal is to understand the data explain the relationship between features and the target better.
• Focus on how to improve, transform, correct, complete, etc. your dataset so that with any model you will have more chance to get better results.
• I do not expect any ML models to be applied to the dataset. I just want to ensure that you have worked on the dataset, understand it, transform it, correct and complete it, and prepare it.
• For each figure, plot, etc. in the presentation, you need to comment on it and explain how and what to do with this information to the dataset. For example, if you have some missing data in some features, tell me how you plan to handle it.
• If you plan to create new features or remove them, tell us why and how.
• If you transform some feature values, explain your need and purpose.
• For any planned action, relate the action with the target task/value (e.g. classification/predicting car prices)
• For feature engineering, at least consider binning, binarize, dummy variables, categoric-to-one-hot encoding, scaling, transformation, etc.
• Please present the material in 10 minutes.
• Each team member is expected to present and know the data and EDA.
• Asking questions to other teams will return you as a bonus grade!

Before the presentation day:
• Presentation should cover the topics given above,
• Select your presenter,
• Rehearse the presentation on Zoom with your team friends,
• Ensure that you can present the material in 10 minutes maximum,
• Be familiar with the Zoom tool to present smoothly,
• Be ready to answer questions!

Good Luck!