Task tracking for MLPA tutorials.

KNN

• Task A - Read data and make predictions pn the wisconsin-breast-cancer dataset using KNN
• Task B - Change K and observe differences
• Task C - KNN using data from sklearn
• Extra - Visualize classifier boundary

Linear regression

• Task A - Simple linear regression
  • Load diabetes ds, plot BMI and y
  • Execute lr given on slides
• Task B - Multiple linear regression
  • LR on boston housing dataset, add features using PolynomialFeatures
  • Scatterplot matrix using pandas scatter_matrix
  • Produce VIF table (variance_inflation_factor)
  • Explore potential multicollinearity
  • Residuals plot using Yellowbrick
• Task C - Lasso regression
  • Plot to compare simple lr, ridge and lasso on slides

Decision Trees

• Task 1A - DT classifier
  • Run dt classifier on breast cancer data from sklearn
  • Find feature importances
  • Print confusion matrix and compare with Week4 (KNN)
• Task 1B - Plotting DT with graphviz
  • Display DT using Graphviz
  • Plot feature importances
• Task 2 - DT regressor
  • Apply DT regressor on sklearn diabetes ds
• Task 3 - Visualizing decision boundaries
  • make_blobs scatterplot
  • make_blobs classifier visualization
• Task 4 - Random forests
  • Classify handwritten numbers using DT+RF
  • Evaulate difference in accuracy between DT and RF
  • Tune n_estimator in RF and explore accuracy
  • Evaluate metrics.classification_report()
  • Change split criteria from Gini to Entropy
  • Apply RandomForestRegressor on Boston housing and compare with previous LR