Neuro-fuzzy Approach - Computational Intelligence in Time Series Forecasting - page 250

s im fzTS .m : N e uro -fuzzy o utp ut vs . A ctua l

1.5

1

0.5

300 400 500 600 700 800 900 1000

tim e

simfzTS.m: Neuro-fuzzy prediction error

0.2

0.1

0

-0 .1

-0 .2

300 400 500 600 700 800 900 1000

tim e

Figure 6.8(b). Performance of a Takagi-Sugeno-type multi-input single-output neuro-fuzzy
network in forecasting the Mackey-Glass chaotic time series. Figure 6.8(a) and Figure 6.8(b)
demonstrate the excellent training and forecasting performance of the Takagi-Sugeno-type
multi-input single-output neuro-fuzzy network respectively for the Mackey-Glass chaotic
time series. It is to be noted that the neuro-fuzzy network considered for this problem has
only four inputs and one output and uses only seven Gaussian membership functions for
(fuzzy) partitioning of input universes of discourse and seven fuzzy rules for neuro-fuzzy
modelling.