On Thursday 19th of Dec. 2019, 09:00 AM, the Ph.D. student Sarmad Mahmud Hadi defended his thesis entitled:
DESIGN AND IMPLEMENTATION OF CLASSIFICATION AND IDENTIFICATION SYSTEM FOR FORENSIC SEQUENCES.
The aim of this work is to design and implement an identification system for forensic DNA based on gel-electrophoresis digital images. Two proposal systems had been designed and implemented, the first one deals with 10 gel-electrophoresis images as input to the neural network (10 profiles), while the second one deals with ten thousand simulated datasets. STR analyses have been done for 10 profiles using the genetic analyzer, where DNA extracted from flinders technology associates (FTA) cards and then sent to the profiling step with 15 loci. Four loci had been selected according to their diversity (non-similarity) to be used as an input to the first proposed system. Gel electrophoresis has been applied to the ten profiles polymerase chain reaction (PCR) product using a marker designed especially only for those four loci. The neural network had been used to learn about DNA features through Gel-electrophoresis images. The dataset used in this thesis is 10 real gel electrophoresis images. The features extracted from the images were fed into the neural network learning about their features testing. The result of the recognition rate was 100%.
The attendance was opened for students and professors interested in the thesis’s subject where the committee examined the student’s knowledge, methodology, and results. At the end of the discussion, the thesis was accepted with minor corrections within a period of one month.