Optimization methods allow larger numbers of points to be specified than do the precision methods, limited only by available computer time and numerical roundoft er ror. Table 5-5 shows a variety of optimization schemes ranging from the mundane (least squares) to the esoteric (fuzzy logic, genetic algorithms). All require a computer-pro- grammed solution. Most can be run on current desktop computers in reasonably short times. Each different optimization approach has advantages and disadvantages in respect to convergence accuracy, reliability, complexity, speed, and computational burden. Convergence often depends on a good choice of initial assumptions ess values) for the linkage parameters. Some methods, if at all, do so to a minimum (only one of many possible solutions), and it may not be the best one for the task.