Please enable JavaScript.

Coggle requires JavaScript to display documents.

Deep Learning (Machine Learning (Tasks (Structured output (NLP Tree),…

- - - - NLP Tree
  - - - Statistical Learning Theory
        
        Data Generating Process:
        The train and test data are generated by a probability distribution over dataset
        
        underfitting:
        Make the training error small
        
        overfitting:
        Make the gap between training and test error small
        
        Capacity
        
        Hypothesis Space
      - Regularization
  - - - Probabilistic
      - Support Vector Machines
        
        Kernel trick
        
        Kernel
        
        polynomial
        
        radial basis function
        
        sigmoid
        
        linear
        
        Trick
      - k -nearest neighbors
      - decision tree
    - - PCA
      - k-means Clustering
  - - - manifold
- - - - affine space
        
        paralleism
        
        ratio of lengths for parallel line segments
        
        displacement vectors (translation)
      - linear combinations in which the sum of the coefficients is 1
      - Affine Transformation
        
        Linear Transformation + translation （平移）
  - - - local estimators
      - Kernel Machine
        
        Support Vector Machine (SVM)
        
        Gaussian processes
        
        Nadaraya-Watson or Parzen windows
        density
        
        Regression estimators
        
        manifold learning
        algorithms
        
        Isomap
        
        LLE
- - - - unbiased estimation
      - asymptotically unbiased
    - - Estimator
      - estimand
        
        quantity of interest
      - estimate
        
        result
    - - Maximum A Posteriori (MAP) Estimation
    - - Cross-entropy
  - - - concave
        ( -f is Convex )
      - local Minimum == global minimum