Modeling and forecasting of a time series data is an integral part of the Data Mining. Sun spot numbers observed on the sun are a good candidate for a time series. A number of linear statistical models are discussed in this paper because Taylor series has similarity with an Auto Regressive model. A new algorithm based on Taylor series expansion and artificial neural network is presented. Based on Taylor series algorithm and ARIMA model, the Sunspot numbers are forecasted and compared.

No abstract

No abstract

In this paper, a novel hybrid model for fitting and forecasting a univariate time series is developed based on ARIMA and HyFIS models. The linear part is fitted using ARIMA model whereas the non-linear residual is fitted using HyFIS model. Clustering technique is used to determine the number of inputs and the membership functions of the HyFIS model. The hybrid model is applied to the wind speed data. The result is analyzed and compared on the basis of standalone ARIMA, standalone HyFIS and for the hybrid ARIMA-HyFIS model.

In this paper, de-noising of satellite attitude and rate data using discrete wavelet transform is presented. The mission objective of any remote sensing satellite is to produce an image. The image is taken on board the spacecraft, and then downloaded using a Radio Frequency link. The processing of the image is done at ground. The attitude quaternion and body rates are part of the down linked data. These data had to be pre-processed to remove the noise. Conventional methods employ taking the Fourier transform and removing the high frequency component through filtering. The de-noising of attitude and rate data is important to make the final data product such as image smoother. General TermsSpacecraft, attitude, body rates.