1 Instructions
• Use python programming language for your implementation.
• Use appropriate approach if you find some attribute is missing in your data.
• Report must contain step-wise description of your implementation and analysis of results. Since data analysis is a crucial task for any machine learning algorithm, report should demonstrate detailed analysis of results and conclusion. It should also clearly mention the steps to run your code.
• Learn the projection matrix for any dimension reduction technique using the train split only. Once the projection matrix has been trained using the train split, use that matrix to reduce the dimension of validation and test splits.
• You can use any python library function to complete the assignment.
2 Dataset:
Download Occupancy Detection Data Set: https://archive.ics.uci.edu/ml/datasets/Occupancy+Detection+
3 Problem statement: Support Vector Machine classifier
In this assignment, you will learn to use several dimensionality reduction techniques to reduce the feature dimension of a data. Then you will train a SVM classifier on the reduced dimension feature space.
1
6. Is there any significant difference between the final test accuracy obtainedfrom Step 3 and Step 5. If so, justify the results with proper reason.
2
