Inaam Ilahi

Freelancer | Python Expert

Lead Backend Developer | Data Engineer | Data Scientist | Technical Project Manager

• Developed optimization algorithms to solve various real-world problems at hand.
  • Developed an algorithm to assign orders to a given fleet based on various calculations. This algorithm increased the order assignment's success rate from 70% (manual assignment) to 90% (algorithmic assignment).
  • Optimized this algorithm to increase speed and reduce complexity to handle more orders.
  • Increased the maximum daily order count from 1000 to 3500.
  • The system I designed and developed is scalable and can handle more than 10,000 orders.
  • Developed scripts that manage these orders and perform updates regularly.
• Designed schema for databases and developed complex REST APIs for multiple projects:
  • Rider On-boarding System (Dashboard | Desktop Application)
  • Order Management System (Dashboard | Desktop Application | Mobile Application)
  • AI Assistant (Mobile Application)
  • Organization Management System (Dashboard | Desktop Application)
  • Payment Management System (Dashboard | Desktop Application)
• Worked on projects in Machine Learning, Deep Learning, and AI:
  • Designing and developing the KMeans algorithm to increase the speed of order assignment.
  • Testing various algorithms on a dataset provided by the client to find patterns in data.
• Developed multiple scripts to crawl and scrap data from numerous websites. Created multiple ETL pipelines for data processing focusing on efficiently and effectively maintaining the data.
  • Extracted doctor's information from a list of websites the client provided and populated the data in a database.
  • Extracted complete data from Trip Advisor and downloaded and saved all the images.
  • Extracted news articles from several websites and populated a single database to search through that data. Developed scripts that continuously go through those websites looking for new articles.
  • Developed a scrapper to crawl through all the websites looking for emails.
  • Developed scripts to crawl and scrap data for numerous other websites as per the client's requirements.
• Developed web automation scripts to automate certain tasks over the internet and generate required results.
• Managed a team of junior to mid-level back-end developers.
  • Deciding the tasks for the team and explaining the client requirements.
  • Managing the tasks of the team on the Jira board.
  • Verifying and approving pull requests.

Research Assistant

I worked under the supervision of Prof. Junaid Qadir. Major part of my research included applying deep reinforcement learning for solving multiple problems.

• Researched the ethics, explainability, and privacy of Natural Language Processing.
• Researched on Incentive-Aware Federated Learning.
• Researched on Deep Image Prior. We performed multiple adversarial attacks on Deep Image Prior algorithms and worked on robustifying them.
• Assisted in teaching the "IHSAN for Muslim Professional" course.
• Researched on applying Deep Reinforcement Learning in networks and its' vulnerabilities and how these models can be robustified.

Graduate Research Fellow and Teaching Assistant

• Researched on Deep Reinforcement Learning in IoT with Dr. Junaid Qadir.
• Researched robust ML models with Dr. Umar Janjua.
• Assisted Sarfraz Raza in teaching two courses, i.e., Computing Fundamentals & Programming, and Object-Oriented Programming.

Engineering Intern

I worked on designing and implementing the human machine interfaces for programmable logic controllers applied in industry. Furthermore, I even helped in the internal factory acceptance testing of PLCs before being deployed in industry.

Project Intern

I learned about the basics and some advanced topics of industrial automation.

Examining Machine Learning for 5G and Beyond through an Adversarial Lens

Abstract
Spurred by the recent advances in deep learning to harness rich information hidden in large volumes of data and to tackle problems that are hard to model/solve (e.g., resource allocation problems), there is currently tremendous excitement in the mobile networks domain around the transformative potential of data-driven AI/ML based network automation, control and analytics for 5G and beyond. In this article, we present a cautionary perspective on the use of AI/ML in the 5G context by highlighting the adversarial dimension spanning multiple types of ML (supervised/unsupervised/RL) and support this through three case studies. We also discuss approaches to mitigate this adversarial ML risk, offer guidelines for evaluating the robustness of ML models, and call attention to issues surrounding ML oriented research in 5G more generally.

Intelligent Resource Allocation in Dense LoRa Networks using Deep Reinforcement Learning (Under Review)

Abstract
The anticipated increase in the count of IoT devices in the coming years motivates the development of efficient algorithms that can help in their effective management while keeping the power consumption low. In this paper, we propose LoRaDRL and provide a detailed performance evaluation. We propose a multi-channel scheme for LoRaDRL. We perform extensive experiments, and our results demonstrate that the proposed algorithm not only significantly improves long-range wide area network (LoRaWAN)'s packet delivery ratio (PDR) but is also able to support mobile end-devices (EDs) while ensuring lower power consumption. Most previous works focus on proposing different MAC protocols for improving the network capacity. We show that through the use of LoRaDRL, we can achieve the same efficiency with ALOHA while moving the complexity from EDs to the gateway thus making the EDs simpler and cheaper. Furthermore, we test the performance of LoRaDRL under large-scale frequency jamming attacks and show its adaptiveness to the changes in the environment. We show that LoRaDRL's output improves the performance of state-of-the-art techniques resulting in some cases an improvement of more than 500% in terms of PDR compared to learning-based techniques.

LoRaDRL: Deep Reinforcement Learning Based Adaptive PHY Layer Transmission Parameter Selection for LoRaWAN

Abstract
The performance of densely-deployed low-power wide-area networks (LPWANs) can significantly deteriorate due to packets collisions, and one of the main reasons for that is the rule-based PHY layer transmission parameters assignment algorithms. LoRaWAN is a leading LPWAN technology where LoRa serves as the physical layer. Here, we propose and evaluate a deep reinforcement learning (DRL)-based PHY layer transmission parameter assignment algorithm for LoRaWAN. Our algorithm ensures fewer collisions and better network performance compared to the existing state-of-the-art PHY layer transmission parameter assignment algorithms for LoRaWAN. Our algorithm outperforms the state of the art learning-based technique achieving up to 500% improvement of PDR in some cases.

Challenges and Countermeasures for Adversarial Attacks on Deep Reinforcement Learning

Abstract
Deep Reinforcement Learning (DRL) has numerous applications in the real world thanks to its outstanding ability in quickly adapting to the surrounding environments. Despite its great advantages, DRL is susceptible to adversarial attacks, which precludes its use in real-life critical systems and applications (e.g., smart grids, traffic controls, and autonomous vehicles) unless its vulnerabilities are addressed and mitigated. Thus, this paper provides a comprehensive survey that discusses emerging attacks in DRL-based systems and the potential countermeasures to defend against these attacks. We first cover some fundamental backgrounds about DRL and present emerging adversarial attacks on machine learning techniques. We then investigate more details of the vulnerabilities that the adversary can exploit to attack DRL along with the state-of-the-art countermeasures to prevent such attacks. Finally, we highlight open issues and research challenges for developing solutions to deal with attacks for DRL-based intelligent systems.

Information Technology University (ITU), Lahore, Pakistan.

My thesis was supervised by Dr. Junaid Qadir. Apart from that, I took the following courses:

• Deep Learning by Dr Mohsen Ali
• Machine Learning by Dr. Ali Ahmed
• Natural Language Processing by Dr. Agha Ali Raza
• Computer Vision by Dr. Mohsen Ali
• Advanced Computer Architecture by Mr. Bilal Hashmi
• Advanced Algorithm Analysis by Dr. Waqas Sultani
• Advanced Operating Systems by Madam. Momina Azam
• Advanced Automata by Dr. Falak Sher

National University of Sciences and Technology (NUST), Islamabad, Pakistan.

My thesis was supervised by Dr. Fahad Mumtaz Malik.