An instructor from the Department of Computer Science receives an international certificate on improving the problem of software Max regression

The lec. Taiba Walaa El-Din Khairi Saeed, Director of the Division of Scientific Affairs and Cultural Relations in the Department of Computer Science at the University of Technology, obtained an international certificate through her participation in the research entitled "Improving the problem of software max retraction and identifying handwritten lines using the Tensor Flow Library" at the World Conference Which was established in Australia under the auspices of the IEEE Foundation and in global cooperation with the Study Group of Charles Sturt University (CSU) in Australia on a voluntary basis.

The aim of the research was to improve the classification accuracy of the current handwritten number systems, and thus improve their efficiency. The proposed system consists of a decision function enhanced by the addition of the "likelihood of bias" function. The function adds negative weights to classes of outputs that have high positive bias and adds positive weight to classes of outputs that have high negative bias in order to neutralize the effect of this high negative bias. Therefore, the Bayesian Classifier function is improved and thus the classification accuracy is improved, which will further improve the performance of the multi-class likelihood classification.

A 5.6% increase in the overall accuracy of classifying handwritten numbers was observed using the revised National Institute of Standards and Technology (MNIST) data set