Machine Learning
Term Project
You task is to perform multiclass classification on the dataset provided:
Minimum requirements:
?
?
?
?
?
?
?
?
?
?
Dataset analysis and report on important statistics
Correlation analysis
Dealing with missing values (if applicable)
Dealing with imbalanced data (if applicable)
Feature transformation/engineering
List of appropriate evaluation measures with justifications
Three Classifiers:
o One of linear classifier (logistic regression or SVMs)
o One of: KNN, Decision Trees, Neural Networks
o One of Ensemble Learning (Random Forest, Adaboost,JProper hyper-parameter tuning based on separate validation set (or crossvalidation)
Error analysis and possible improvements
Final results on the test set
Other possible ideas to try (as examples):
?
?
?
?
?
?
?
?
More than 3 classifiers and comparison
Investigate the concept of margins
Feature importance, feature selection, dimensionality reduction
Investigate different feature scaling techniques
Investigate different techniques for encoding categorical values.
Clustering the data in K clusters (K= number of classes) and compare the labels
Interpreting the learned models (for example by examining the weights of a linear
model or by constructing decision rules from the learnt decision tree)
‰mportant Notes:
1. All the documents (code and report) should be submitted in Jupyter notebooks.
2. Your work will be checked with appropriate plagiarism detection tools like iThenticate.
3. You can work as a team of 2 members.
4. Best performing team gets a bonus.

Purchase answer to see full
attachment