Orah Richard | Federal polytechnic Idah (original) (raw)

Orah  Richard

Research on Machine Learning Task and Software DeveloperPhone: 08051804594, 08067613144Address: Behind Fosla Hotel, Idah Kogi State NigerialessEmail:orahseun@gmail.com
Phone: 08051804594, 08067613144
Address: Behind Fosla Hotel, Idah Kogi State Nigeria

less

Related Authors

GJESRM Journal

IJERA Journal

IJERA Journal

International Journal of Engineering Research and Applications (IJERA)

Waseem Shahzad

Abidemi Adeniyi

Uploads

Papers by Orah Richard

Research paper thumbnail of FORECASTING INFLATION VARIABLES USING ARTIFICIAL NEURAL NETWORK TECHNIQUES

The analysis of monthly Inflation constitutes one of the major economic problems in emerging mark... more The analysis of monthly Inflation constitutes one of the major economic problems in emerging market economies that requires monetary authorities to elaborate tools and policies to prevent high volatility in prices and long periods of inflation. This work modeled and forecast monthly inflation rate of Nigerian by using neural networks on the evaluation of set of variables. The data set used for estimating the models are obtained from central bank statistical data base for the period of 2000 to 2017. The neural network models Back propagation neural network model or Back propagation Network model (BPN model) was used in this study in forecasting inflation rate. The results of the neural network models under static with the traditional econometric model and the results shows that the performance is better when compare it with the traditional econometric model in forecasting the inflation rate

Research paper thumbnail of report.doc AUTORSHIP ATTRIBUTION  USING NEURAL NETWORK FINAL YEAR PROJECT

Research paper thumbnail of AUTORSHIP ATTRIBUTION USING NEURAL NETWORK FINAL YEAR PROJECT

This project work presents a model for authorship attribution that employs the neural network lea... more This project work presents a model for authorship attribution that employs the neural network learning methodology. Stylometric features were extracted from texts written by three prominent literary writers of Nigerian origin. The models were learned on a dataset consisting 422 samples from Chinua Achebe, 255 samples from Ola Rotimi and 395 samples from Wole Soyinka. Our experimental results indicate classification accuracies of 85%, 72%, and 80% for Chinua Achebe, Ola Rotimi and Wole Soyinka respectively in the best case. From the analysis of our results, we see that neural network copes well with undiscretized data set as shown in the ROC curves obtained from each experiment but we add quickly that these results are not of sufficient accuracy to actually be of much practical use. Hence the effort of future research is to target discretization as a necessary step in the model estimation.
Keyword: Stylometry, Machine Learning and classifier

Research paper thumbnail of FORECASTING INFLATION VARIABLES USING ARTIFICIAL NEURAL NETWORK TECHNIQUES

The analysis of monthly Inflation constitutes one of the major economic problems in emerging mark... more The analysis of monthly Inflation constitutes one of the major economic problems in emerging market economies that requires monetary authorities to elaborate tools and policies to prevent high volatility in prices and long periods of inflation. This work modeled and forecast monthly inflation rate of Nigerian by using neural networks on the evaluation of set of variables. The data set used for estimating the models are obtained from central bank statistical data base for the period of 2000 to 2017. The neural network models Back propagation neural network model or Back propagation Network model (BPN model) was used in this study in forecasting inflation rate. The results of the neural network models under static with the traditional econometric model and the results shows that the performance is better when compare it with the traditional econometric model in forecasting the inflation rate

Research paper thumbnail of report.doc AUTORSHIP ATTRIBUTION  USING NEURAL NETWORK FINAL YEAR PROJECT

Research paper thumbnail of AUTORSHIP ATTRIBUTION USING NEURAL NETWORK FINAL YEAR PROJECT

This project work presents a model for authorship attribution that employs the neural network lea... more This project work presents a model for authorship attribution that employs the neural network learning methodology. Stylometric features were extracted from texts written by three prominent literary writers of Nigerian origin. The models were learned on a dataset consisting 422 samples from Chinua Achebe, 255 samples from Ola Rotimi and 395 samples from Wole Soyinka. Our experimental results indicate classification accuracies of 85%, 72%, and 80% for Chinua Achebe, Ola Rotimi and Wole Soyinka respectively in the best case. From the analysis of our results, we see that neural network copes well with undiscretized data set as shown in the ROC curves obtained from each experiment but we add quickly that these results are not of sufficient accuracy to actually be of much practical use. Hence the effort of future research is to target discretization as a necessary step in the model estimation.
Keyword: Stylometry, Machine Learning and classifier

Log In