Difference between revisions of "Terms"

Revision as of 20:02, 21 June 2015

Artificial Intelligence Nouns

@@Home -> ArtificialIntelligenceDictionary -> terms

Common Entities

• action
  • 1. continuous action
• action selection strategy
  • 1. confidence based exploration (Thrun, 1999)
    2. directed exploration
    3. eps.-greedy selection
    4. error-based directed exploration
    5. frequency-based directed exploration
    6. optimism in the face of uncertainty
    7. recency-based directed exploration (Sutton, 1990)
    8. tabu search (Abramson and Wechsler, 2003)
• activation function
  • 1. hyperbolic tangent activation function
    2. linear activation function
    3. logistic function
    4. monotonic activation function
    5. normal sigmoid function
    6. periodic activation function
    7. sigmoid function
    8. symmetric sigmoid function
    9. symmetric sinus activation function
    10. threshold activation function
• agent
  • 1. autonomous agent
• artificial intelligence
• back-propagation drawbacks
  • 1. local minima problem
    2. moving target problem
    3. step-size problem
• belief nets
  • 1. directed belief nets
    2. sigmoid belief nets
• binary codes
• cause
• cascade correlation architecture (Fahlman and Lebiere, 1990)
• conditional random fields
• connection
  • 1. autoregressive connections
    2. input connections
    3. lateral connection
    4. output connections
    5. short-cut connections
    6. symmetric connections
    7. temporal connections
    8. trainable connections
• containment function
• damping
• dataset
  • 1. labeled data
    2. noise-free data
    3. sample
    4. sequential data
    5. test set
    6. training example
    7. training patterns
    8. training data-set
    9. unbiased example
    10. unlabeled data
    11. validation data-set
• dimensionality reduction
  • 1. non-linear dimensionality reduction
• discount rate
• directed model
• distributed representations
• domain-specific kernel
• dynamic programming
• eligibility traces
  • • 1. replacing eligibility traces
• energy of joint configuration
• environment
  • 1. stationary environment
• epoch
• error value
  • 1. mean square error (MSE)
• experience value
  • 1. discounted future experience
    2. immediate experience value
• experience value function
• factorial distribution
• feature
• generative model
• generalization
• goal state
• gradient
• greedy strategy
• inference
• layer
  • 1. input layer
    2. hidden layer
    3. layer of features
    4. output layer
• learning rate
• likelihood
• local optima (for neural network)
• log likelihood
• log probability
• misclassification rate
• neural networks
  • 1. artificial neural network (ANN)
    2. cascading neural networks
    3. convolutional multilayer neural networks
    4. counterpropagation network
    5. deep neural networks
    6. feedforward networks
    7. fully connected neural network
    8. functional-link neural networks
    9. general regression neural network
    10. higher order networks
    11. multilayer feedforward artificial neural networks
    12. multilayer neural networks
    13. probabilistic neural network
    14. real-time recurrent learning networks
    15. recurrent backpropagation networks
    16. recurrent neural networks
• neuron
  • 1. bias neuron
    2. binary neurons
    3. candidate neuron
    4. hidden neuron
    5. mean-field logistic unit
    6. output neuron
• node (in the network)
  • 1. leaf node (in the network)
    2. unit
• noise (in the data)
• objective function
• online inference
• output
  • 1. actual output
    2. desired output
• over-fitting
• partial derivative
• policy
  • 1. deterministic policy function
    2. optimal policy
    3. optimal deterministic policy
    4. stochastic policy function
• posterior distribution
  • 1. aggregated posterior distribution
• precision-recall curves
• prior
  • 1. complementary prior
• probability
• probability density models
• profit function
• reward
  • 1. cumulative reward
    2. discounted future reward
    3. future reward
    4. immediate reward
    5. longterm reward
    6. short-term reward
• reward value function
• root mean squared error
• second order statistics
• selective attention approach
• sensory input
• shallow models
• slackness of the bound
• sloppy top-down specification
• softmax function
• state
  • 1. after-state
    2. continuous state
• state-action space
• stop function
• structure (in the data)
• training curve
• value function
  • 1. action-value function
    2. state-value function
• variable (for neural network)
  • 1. circular variables
    2. stochastic variable
• weights
  • 1. frozen weights
    2. initial weights
    3. lateral weight

Named Entities

• Adaline
• ARTMAP Neural Networks
  • 1. Fuzzy ARTMAP
    2. Gaussian ARTMAP
• Bellman Optimality Equation (Sutton and Barto, 1998)
• Bernoulli Variables
• Bidirectional Associative Memory (BAM)
• Boltzmann Machine
  • 1. Conditional RBM model
    2. Restricted Boltzmann Machine (RBM)
    3. Semi-restricted Boltzmann Machines
    4. Temporal RBM
• Boltzmann-Gibbs Selection
• Deep Belief Nets
  • 1. Deep Autoencoders
• Dynamic Bayes Nets
• Elman Neural Networks
• Finite Impulse Response (FIR) filter
• Gaussian Processes
• Gaussian Unit
• Hebbian Theory
• Hidden Markov Models (HMM)
• Hopfield Net
• Jordan Neural Network
• Long Short-Term Memory (LSTM) Recurrent Network
• Markov Decision Process (MDP)
• Markov Environment
• Markov Property
• Markov State
• Max-Boltzmann Selection
• MNIST Test Set
• MRF
  • 1. MRF-MBNN
• Neocognitron
• Perceptron
• RBF Networks
• Support Vector Machine (SVM)
• T-step policy
• T-step return
• TF-IFD
• Threshold Logical Units (TLU) Network
• Time Delay Neural Network (TDNN)
• UNI-SNE method