ISE 468 KfUPM Machine Learning and Data Analytics Paper

Homework #1ISE 468 Introduction to Machine Learning and Data Analytics, Term: 221
In this team assignment, answer each item with your own words; you can use images, block diagrams,
formulas from other resources (online or book) with their references. Exact copies of definitions from
those resources will not be accepted. For each item, provide (at least) one paragraph explanation and
elaborate your answers with figures, images, block diagrams and formulas if necessary. In your answers
for each item, make sure you provide references for the resources you use.
Item 1. Explain Bias-Variance Tradeoff in predictive models.
Item 2. Explain Class Imbalance problem in classification modeling.
Item 3. Explain Overfitting and Underfitting in predictive modeling.
Item 4. Explain the Variable Transformation in data preprocessing
Item 5. Explain pre-scaling methods used in data preprocessing and give at least 3
methods for scaling; their definitions, formulas, pros and cons
Item 6. Explain the Distance metrics used in similarity of data instances and give at least 3 examples
for distance metrics; their definitions, formulas, pros and cons
Item 7. Explain the Validation Methods in evaluating predictive models and give at least 3
examples for validation methods; their definitions, formulas, pros and cons
Item 8. Explain the role of Loss functions in Classification models and give at least 3 examples for
classification loss functions; their definitions, formulas, pros and cons
Item 9. Explain the Error metrics used in evaluating Classification models and give at least 3 examples
for classification error metrics; their definitions, formulas, pros and cons
Item 10. Explain Ensemble modeling methods (bagging, boosting and random forest); their definitions,
formulas, pros and cons
Item 11. Explain Gini Index, Entropy, Information Gain, and Gain Ratio used in Decision Tree modeling
Item 12. Explain the clustering performance metrics used in evaluating clustering models and give at
least 3 metrics; their definitions, formulas, pros and cons
Item 13. Explain backpropagation operation in training neural networks
Item 14. Explain Mercer’s Theorem and Kernel Trick in Support Vector Machines
Item 15. Give 3 examples for kernel functions used in Support Vector Machines: their definitions,
formulas, pros and cons.
References (you can use other resources as well):
•
•
•
•
•
•
Bias–variance tradeoff – Wikipedia
Gentle Introduction to the Bias-Variance Trade-Off in Machine Learning
(machinelearningmastery.com)
A Gentle Introduction to Imbalanced Classification (machinelearningmastery.com)
Bias Variance Tradeoff | What is Bias and Variance (analyticsvidhya.com)
6. Dataset transformations — scikit-learn 1.0.1 documentation
6.3. Preprocessing data — scikit-learn 1.0.1 documentation
sklearn.preprocessing.FunctionTransformer — scikit-learn 1.0.1 documentation
sklearn.neighbors.DistanceMetric — scikit-learn 1.0.1 documentation
3.1. Cross-validation: evaluating estimator performance — scikit-learn 1.0.1 documentation
Loss functions for classification – Wikipedia
Common Loss functions in machine learning for Classification model | by Sushant Kumar |
Analytics Vidhya | Medium
3.3. Metrics and scoring: quantifying the quality of predictions — scikit-learn 1.0.1
documentation
1.11. Ensemble methods — scikit-learn 0.16.1 documentation
Ensemble Machine Learning Algorithms in Python with scikit-learn
(machinelearningmastery.com)
Backpropagation – Wikipedia
Mercer’s theorem – Wikipedia
lec_10_09_2013.pdf (princeton.edu)
http://scikit-learn.org/stable/modules/clustering.html
•
http://scikit-learn.org/stable/modules/clustering.html#clustering-performance-evaluation
•
•
•
•
•
•
•
•
•
•
•
•
•