The evolution of the KSE-100 index
Bilal Ahmed Usmani
Friday 11th November, 2011 16:00-17:00 Mathematics Building, room 516
This study discusses the evolution of KSE-100 index returns series R as a dynamical system. To do this, First, we perform nonlinear time-series analysis of R and attempt to compute correlation dimension, dc. Our results show that estimation of dc for the series R is not possible. Secondly, we construct feedforward neural network models of R with backpropagation training. Lastly, a Comparison of Neural Network models with conventional ARMA/ARIMA models is performed. Neural networks are found to be applicable in those cases when nonlinear time-series analysis is at failure.