Machine Learning in a Random Forest

Celebrating the KSML Science Comic app surpassing 10,000 downloads on Google Play, we now officially release “Machine Learning in a Random Forest,” the Ultimate Edition.

Module 1: model, evaluation, overfitting, train-test-split

1. The model that’s not a girl!

2. Supervised learning: who’s supervising the forest?

3. Unsupervised learning helps detect shady people

4. Parameters and Loss function

5. Simple Linear Regression Review: Sunlight & Selfie

6. Evaluation measure: MSE versus MAE, RMSE

7. model generalization overfitting, underfitting

8. train-test split

a crab’s feeling of underfitting

Module 2: Regression

Multiple linear regression

Residual plot for model diagnostic

Non-constant variance in linear regression: a duck’s mood swing problem

Sum of Squares & coefficients of determination with Python & R codes

Limitation of R-squared, Adjusted R squared

Module 3: Feature Selection & Regularization

Feature selection & Model Selection

Forward feature selection: a step by step example

Backward feature selection + example

Stepwise Feature Selection +example

Comparing forward, backward, stepwise feature selection

Review: Maximum Likelihood Estimation

Lasso & Ridge regression: method & codes

ElasticNet Regression: Method & Codes

Module 4: Hyper parameter tuning

Grid search and train-validation-test split for hyperparameter tuning-intro

Hyperparameter tuning by train-validation-test split – process & example

Module 5: Classification

What’s classification

K-Nearest Neighbors (KNN): an introduction

KNN classification: practical notices & implementation using Python & R

Logistic Regression: method + Python & R codes

Logistic regression with L1 or L2 penalty (advanced)

Support Vect or Machine + Python & R Codes

Kernel tricks, SVM properties & kernel choice

What’s a decision tree? Where should the squirrels store her acorns?

Module 6: Unsupervised learning

Dimension reduction and Principal Component Analysis (PCA)

Feature selection versus dimensionality reduction

K-Means Clustering

Module 7: Advanced topics

Bootstrap: introductory comics & quizzes

Bagging & Random Forest: intro & quizzes

error: Content is protected !!