Wednesday, June 15, 2016

Learning and Adaptation


  • Supervised Learning
  • Unsupervised Learning
  • Reinforcement Learning

The Sub-problems of Pattern Classification


  1. Feature Extraction
  2. Noise
  3. Overfitting
  4. Model Selection
  5. Prior Knowledge
  6. Missing Features
  7. Mereology
  8. Segmentation
  9. Context
  10. Invariances
  11. Evidence Pooling
  12. Costs and Risks
  13. Computational Complexity

The "eight hats" of data visualization design


  1. The initiator is the leader, the person who is seeking a solution to the task as per the brief or self-initiated curiosity. 
  2. The data scientist is characterized as the data miner, wearing the miner's hat.
  3. The journalist is the storyteller, the person who establishes the narrative approach to the visualization's problem context.
  4. The computer scientist is the executor, the person who brings the project alive. 
  5. The designer is the creative, the one, who, in harmony with the computer scientist, will deliver the solution. 
  6. The cognitive scientist is the thinker in terms of appreciating the science behind the effectiveness of the technical and designed solutions. 
  7. The communicator is, naturally, concerned with the communication side of the project.
  8. The project manager is the person who does much to pick up many of the unpopular duties to help bring the whole project together.