Machine Learning Algorithms
Table of contents
kNN
k Nearest Neighbor classifier
How To
For the query data point, find the k closest training data points,
and predict using their labels.Hyper Parameter
k, metricIssues
- time complexity : $O(1)$ for training, $O(n)$ for predicting
- curse of dimensionality (exponential)
- In computer vision, pixel-level nearness doesn’t represent high-level similarity.
Linear Classifier
by affine transformation
SVM
Support Vector Machine
Tree
Ensemble
GMM
K-means
clustering
PCA
Principal Component Analysis
SGD
Stochastic Gradient Descent
CNN
Convolutional Neural Network
RNN
Recurrent Neural Network