Neural Knapsack: A Neural Network Based Solver for the Knapsack Problem
IEEE Access 2020 Hazem AA Nomer, Khalid Abdulaziz Alnowibet, Ashraf Elsayed, Ali Wagdy Mohamed
Hi, my name is Hazem Nomer. I am a machine learning researcher interseted in areas at the intersection of optimization and deep learning (aka representation learning). Throughout my carrer, I have been intersted in non-convential and challeging ideas in deep learning (as solving mathematical problems with neural networks). I even built a one myself NeuralKnapsack: a neural solver for the Knapsack problem. I am "jack of all trades, master of none" as I have been involved in serveral projects in computer vision, natural langauge processing, combinatorial optimization and reinforcment learning. Currenlty, I am machine learning Researcher at Siemens EDA (Software Digital Industries) working on hot-spot detection problem in lithography. I am enrolled in the PhD program (on leave) at ITCS school, Nile University.
Publications
GSK-RL: Adaptive Gaining-sharing Knowledge algorithm using Reinforcement Learning
NILES 2021 Hazem AA Nomer, Ali Wagdy Mohamed, Ahmed H Yousef
Code will be avaliable soon
Projects
Deep Algorithmic Trading
Stocks returns prediction using deep learning
Mouse RNN
This simple module tracks your mouse movements and predict where your mouse is going next
Stack LSTM
Teaching LSTM and GRU (Gated Recurrent Units) to act as Stacks.
Recursive Neural Networks
Due to many theoretical implications, recursive neural networks are believed to be powerful models. In addition, recursive networks learn high-level representations from explicit inputs thus, it is successful in many deep learning tasks where the input is a structure
Mars Scout
(Role: C# Developer) 3D game simulates life on Mars Lava tubes. Won in NASA SPACE APPS hackathon.