Outline

1. Background
   • "linear" problems
   • types of computations
   
   
2. Sources of computational inaccuracy
   • floating point representation of numbers
   • floating point arithmetic
   • particular types of floating point errors
   
   
3. Selection of datasets
   • sources of candidate datasets
   • "real" versus "constructed" datasets
   • levels of computational difficulty
   
   
4. Computation of error-free results
   • extended precision computations
   • sample algorithms and output
   • checks on computational results