Papers by Mariem Bouhadda
The purpose of this project is to provide a tool that uses breast cancer Histopatho-
logical imag... more The purpose of this project is to provide a tool that uses breast cancer Histopatho-
logical images to predict cancer type by classifying it into benign and malignant tumors.
The stock market is constantly changing with uncertainties. That's why the analysis of certain ti... more The stock market is constantly changing with uncertainties. That's why the analysis of certain time series from the economic and financial world shows specific characteristics that are not theoretically taken
into account in ARIMA modeling Box Jenkins. The first characteristic concerned is the
dependence of the conditional variance of time, in other words, we are concerned by the heteroscedastic behavior of the variance. The second
characteristic concerns the probabilistic distribution of the series (the appearance of shocks is
not compatible with the normal law). Then for a more realistic modeling of these series,
ARCH / GARCH models are the most useful.
In this project my co-worker and I :
► We have studied the ARCH and GARCH models from a theoretical point ( properties and conditions
to be fulfilled by the models).
► We applied them to one of the indices (CAC40 ) using EVIEWS.
► We applied 3 different forecasting models to predict the conditional variance of CAC 40 stock indices.
► We finally compared the forecasts using multiple statistical tests.
Uploads
Papers by Mariem Bouhadda
logical images to predict cancer type by classifying it into benign and malignant tumors.
into account in ARIMA modeling Box Jenkins. The first characteristic concerned is the
dependence of the conditional variance of time, in other words, we are concerned by the heteroscedastic behavior of the variance. The second
characteristic concerns the probabilistic distribution of the series (the appearance of shocks is
not compatible with the normal law). Then for a more realistic modeling of these series,
ARCH / GARCH models are the most useful.
In this project my co-worker and I :
► We have studied the ARCH and GARCH models from a theoretical point ( properties and conditions
to be fulfilled by the models).
► We applied them to one of the indices (CAC40 ) using EVIEWS.
► We applied 3 different forecasting models to predict the conditional variance of CAC 40 stock indices.
► We finally compared the forecasts using multiple statistical tests.
logical images to predict cancer type by classifying it into benign and malignant tumors.
into account in ARIMA modeling Box Jenkins. The first characteristic concerned is the
dependence of the conditional variance of time, in other words, we are concerned by the heteroscedastic behavior of the variance. The second
characteristic concerns the probabilistic distribution of the series (the appearance of shocks is
not compatible with the normal law). Then for a more realistic modeling of these series,
ARCH / GARCH models are the most useful.
In this project my co-worker and I :
► We have studied the ARCH and GARCH models from a theoretical point ( properties and conditions
to be fulfilled by the models).
► We applied them to one of the indices (CAC40 ) using EVIEWS.
► We applied 3 different forecasting models to predict the conditional variance of CAC 40 stock indices.
► We finally compared the forecasts using multiple statistical tests.