Clustering & Classification
Detect and predict groups in your data to understand differences and make predictions.
Similarity
Forming Clusters
K-Means Algorithm
Interpreting Clusters
Initial Guesses
Inertia
Total Inertia
Optimization
Number of Clusters
More Than Two Variables
Magnitudes
Min-Max Scaling
Outliers
Classification
Misclassified Data
Building a Simple Classifier
Probabilistic Classification
Logistic Regression
Pixels
Weights
Heat Maps
Bias
Probabilities From the Sigmoid
Log Likelihood
Image Classifier Optimization
Multi-class Regression
The Softmax Function
Up next
Regression
Leverage the power of correlation and regression to understand relationships and make predictions with data.
Course description
This course introduces k-means clustering and logistic regression, which are used to uncover natural groupings in data and to model the probability of membership in a set of classes, respectively. It explores two applications of these methods, using k-means clustering to segment observations into meaningful clusters and logistic regression to classify instances and predict the likelihood of specific categories. Lessons cover choosing the optimal number of clusters, comparing the inertia of different clusterings, and minimizing likelihood with stochastic gradient descent. By the end, you'll understand how and when to deploy these methods and how to derive insights from their results.
Topics covered
- k-means clustering
- scatterplots
- inertia
- climate data
- min-max scaling
- outliers
- logistic regression
- classification
- likelihood
- decision boundary
- stochastic gradient descent
- binary random variables
- probability