Machine Learning

Machine Learning

1. Supervised Learning

   • The computer is presented with example INPUTS AND their desired OUTPUTS, given by a "teacher", and the goal is to learn a general rule that maps inputs to outputs.
   • Clasification
     • The output is element of a set (true/false, categories, etc).
     • Neural Network
       • Universal function approximator.
       • Recurrent
       • Convolutional
         • Used for image classification. Based on mathematical process known as "convolution". Non-linear. Layers: - INPUT - Convolution - Activation - Pooling - Fully connected - OUTPUT CONVOLUTION - breaks into features OUTPUT - node per label, for each incoming feature the nodes receive points, then the node with more points is the result
       • Feed-forward
     • Bayes Classifier
     • Suport Vector Machine
     • Decision Trees
   • Regression
     • The output is a real number.

2. Unsupervised Learning == Data Mining

   • Training set provides only inputs, leaving it on its own to find structure in its input.
   • Clustering
     • k-means
     • mixture models
   • Anomaly detection
   • Hebbian
   • Latent Vars

3. Reinforcement Learning

   • Reinforcement learning is concerned with the problem of finding suitable actions to take in a given situation in order to maximize a reward. A computer program interacts with a dynamic environment in which it must perform a certain goal (such as driving a vehicle or playing a game against an opponent). The program is provided feedback in terms of rewards and punishments as it navigates its problem space.