🚇 Operations Research with ML (OR+ML) Wing
Traditional optimization techniques in Industrial and Production Engineering (IPE) provide a robust foundation for solving complex problems and challenges of modern supply chains and manufacturing systems. By integrating machine learning (ML) with optimization, we can unlock new possibilities for enhancing efficiency, decision-making, and process improvement across various domains. CIOL focuses on this intersection, leveraging the strengths of both fields for groundbreaking research. Our goal :
- Develop ML-powered algorithms that efficiently solve complex optimization problems within the constraints of intricate industrial settings.
- Use ML to analyze industrial data, build better optimization models, and improve decision-making under uncertainty, leading to increased profitability.
- Fuse ML models with traditional optimization techniques to create powerful hybrid methods that maximize the strengths of both approaches.
- Develop AI-powered forecasting models to optimize inventory management, production scheduling, and resource allocation in supply chains.
- Create AI-driven risk assessment tools to proactively identify and mitigate disruptions within complex supply chain networks.
- Utilize graph neural networks to optimize transportation routes, improve network resilience, and analyze the inter-connectivity of supply chains.
Wing Members
Publications
-
Neural Control System for Continuous Glucose Monitoring and Maintenance
Azmine Toushik Wasi ICLR'24 Tiny Papers â–ª Accepted â–ª CO â–ª Medical AI [OpenReview] â–ª [arXiv]
- SupplyGraph: A Benchmark Dataset for Supply Chain Planning using Graph Neural Networks
Azmine Toushik Wasi, MD Shafikul Islam Sohan, Adipto Raihan Akib AAAI'24 Graphs and more Complex structures for Learning and Reasoning Workshop (Full Paper) â–ª Accepted â–ª GNN â–ª Supply Chains â–ª Datasets and Benchmarks [Paper Site] â–ª [arXiv] â–ª [GitHub] â–ª [Kaggle]
- Optimizing Inventory Routing: A Decision-Focused Learning Approach using Neural Networks
MD Shafikul Islam Sohan, Azmine Toushik Wasi NeurIPS'23 New in ML Workshop (Extended Abstracts) â–ª Accepted â–ª CO â–ª OR-ML â–ª Supply Chains [Paper Site] â–ª [arXiv] â–ª [OpenReview]
-
Graph Neural Networks in Supply Chain Optimization: Concepts, Perspectives, Dataset and Benchmarks
Azmine Toushik Wasi, MD Shafikul Islam, Adipto Raihan Akib, and Mahathir Mohammad Bappy In Review â–ª GNN â–ª Supply Chain
-
Presecriptive Analytics: A Review in the Landscape of Machine Learning
In Progress â–ª
AML â–ª
SCM