2018: Today I learned

This page is used to track my academic achievements in 2018, starting from 24 Jan.


  • 29
    • doi:10.1088/1741-2560/6/4/046002
    • CCA Algorithm (https://eprints.soton.ac.uk/259225/1/tech_report03.pdf)
      • Showing correlations between 2 matrices
      • Find a vector $w_x$ and $w_y$ for $X$ and $Y$ to be projected on. Optimise both $w_x$ and $w_y$ so that the correlation is maximised.
  •  28
    • Introduction to Brain Controlled Interface
  • 27
    • Ensemble Learning
      • A combination of small learning rules
  • 26
    • Neural networks
      • Perceptron as a Linear Separable Model
      • Limitations of a single perceptron (XOR)
      • Activation function
      • Gradient descents
      • Backpropagation
    • Fundamental concepts on k-NN
  • 25
    • Regression and Classification
  • 24
    • Decision tree ID3 algorithm (top down learning, from the most seperatable key)
    • ID3 bias
    • Tree pruning (training to the end and removing the nodes little by little in order to reduce overfit)
    • Proofing Linear Regression fit with linear algebra (hmm, interesting…)
    • Cross validation training (separate training data into folds, and choose one fold as pseudo-test set.)