LIMITATIONS

The main disadvantage of both on-line and off-line variants is their computational complexity. The current off-line implementation is clearly slower than conventional standard techniques, by about three orders of magnitude (but no attempt was made to optimize the code with respect to speed). And the complexity of the on-line method is even worse (the exact slow-down factor depends on the precise nature of the learning protocol, of course). For this reason, especially the promising on-line variants can be recommended only if special neural net hardware is available. Note, however, that there are many commercial data compression applications which rely on specialized electronic chips.