Machine Learning

Principal Component Analysis (PCA) and t-Distributed Stochastic Neighbor Embedding (t-SNE) are all efficient ways to transform the data points in high dimesion to the corresponding feature points in low dimension without losing the principal feature.

In machine learning, the perceptron is an algorithm for supervised learning of binary classfiers. The perceptron algorithm was invented 1958 at the Cornell Aeronautical lab by Frank Rosenblatt. Perceptron model For input space (featured space) $\mathcal{X} = {{\mathbf{R}}^{n}}$ and output space $\mathcal{Y} =\{-1, +1\}$, the perceptron model can be built by: $$f(x) = sign(\omega \centerdot x+b)$$ $$sign(x) = \begin{cases} +1 & \text{if } x \ge 0,\\ -1 & \text{if } x < 0.