Search Results for author: Sourav Medya

Found 21 papers, 10 papers with code

Game-theoretic Counterfactual Explanation for Graph Neural Networks

no code implementations8 Feb 2024 Chirag Chhablani, Sarthak Jain, Akshay Channesh, Ian A. Kash, Sourav Medya

Our results reveals that computing Banzhaf values requires lower sample complexity in identifying the counterfactual explanations compared to other popular methods such as computing Shapley values.

counterfactual Counterfactual Explanation +2

VeriBug: An Attention-based Framework for Bug-Localization in Hardware Designs

no code implementations17 Jan 2024 Giuseppe Stracquadanio, Sourav Medya, Stefano Quer, Debjit Pal

Then, it assigns an importance score to each operand in a design statement and uses that score for generating explanations for failures.

DGCLUSTER: A Neural Framework for Attributed Graph Clustering via Modularity Maximization

1 code implementation20 Dec 2023 Aritra Bhowmick, Mert Kosan, Zexi Huang, Ambuj Singh, Sourav Medya

Graph clustering is a fundamental and challenging task in the field of graph mining where the objective is to group the nodes into clusters taking into consideration the topology of the graph.

Clustering Graph Clustering +2

COMBHelper: A Neural Approach to Reduce Search Space for Graph Combinatorial Problems

1 code implementation14 Dec 2023 Hao Tian, Sourav Medya, Wei Ye

Combinatorial Optimization (CO) problems over graphs appear routinely in many applications such as in optimizing traffic, viral marketing in social networks, and matching for job allocation.

Combinatorial Optimization Knowledge Distillation +1

NeuroCUT: A Neural Approach for Robust Graph Partitioning

no code implementations18 Oct 2023 Rishi Shah, Krishnanshu Jain, Sahil Manchanda, Sourav Medya, Sayan Ranu

Second, we decouple the parameter space and the partition count making NeuroCUT inductive to any unseen number of partition, which is provided at query time.

graph partitioning

GNNX-BENCH: Unravelling the Utility of Perturbation-based GNN Explainers through In-depth Benchmarking

no code implementations3 Oct 2023 Mert Kosan, Samidha Verma, Burouj Armgaan, Khushbu Pahwa, Ambuj Singh, Sourav Medya, Sayan Ranu

Motivated by this need, we present a benchmarking study on perturbation-based explainability methods for GNNs, aiming to systematically evaluate and compare a wide range of explainability techniques.

Benchmarking counterfactual

Empowering Counterfactual Reasoning over Graph Neural Networks through Inductivity

1 code implementation7 Jun 2023 Samidha Verma, Burouj Armgaan, Sourav Medya, Sayan Ranu

Graph neural networks (GNNs) have various practical applications, such as drug discovery, recommendation engines, and chip design.

counterfactual Counterfactual Explanation +2

A Survey on Explainability of Graph Neural Networks

no code implementations2 Jun 2023 Jaykumar Kakkad, Jaspal Jannu, Kartik Sharma, Charu Aggarwal, Sourav Medya

Graph neural networks (GNNs) are powerful graph-based deep-learning models that have gained significant attention and demonstrated remarkable performance in various domains, including natural language processing, drug discovery, and recommendation systems.

Drug Discovery Recommendation Systems

Global Counterfactual Explainer for Graph Neural Networks

1 code implementation21 Oct 2022 Mert Kosan, Zexi Huang, Sourav Medya, Sayan Ranu, Ambuj Singh

One way to address this is counterfactual reasoning where the objective is to change the GNN prediction by minimal changes in the input graph.

counterfactual Counterfactual Explanation +2

Incorporating Heterophily into Graph Neural Networks for Graph Classification

1 code implementation15 Mar 2022 Wei Ye, Jiayi Yang, Sourav Medya, Ambuj Singh

Graph neural networks (GNNs) often assume strong homophily in graphs, seldom considering heterophily which means connected nodes tend to have different class labels and dissimilar features.

Graph Classification

An Exploratory Study of Stock Price Movements from Earnings Calls

no code implementations31 Jan 2022 Sourav Medya, Mohammad Rasoolinejad, Yang Yang, Brian Uzzi

Third, the semantic features of transcripts are more predictive of stock price movements than sales and earnings per share, i. e., traditional hard data in most of the cases.

Task and Model Agnostic Adversarial Attack on Graph Neural Networks

1 code implementation25 Dec 2021 Kartik Sharma, Samidha Verma, Sourav Medya, Arnab Bhattacharya, Sayan Ranu

In this work, we study this problem and show that GNNs remain vulnerable even when the downstream task and model are unknown.

Adversarial Attack Q-Learning

Event Detection on Dynamic Graphs

1 code implementation23 Oct 2021 Mert Kosan, Arlei Silva, Sourav Medya, Brian Uzzi, Ambuj Singh

In this paper, we propose DyGED, a simple yet novel deep learning model for event detection on dynamic graphs.

Decision Making Event Detection

Meta-Learning with Graph Neural Networks: Methods and Applications

no code implementations27 Feb 2021 Debmalya Mandal, Sourav Medya, Brian Uzzi, Charu Aggarwal

Graph Neural Networks (GNNs), a generalization of deep neural networks on graph data have been widely used in various domains, ranging from drug discovery to recommender systems.

Drug Discovery Meta-Learning +1

Investigating Ground-level Ozone Formation: A Case Study in Taiwan

no code implementations18 Dec 2020 Yu-Wen Chen, Sourav Medya, Yi-Chun Chen

In this paper, we aim to identify and understand the impact of various factors on O3 formation and predict the O3 concentrations under different pollution-reduced and climate change scenarios.

Manipulating Node Similarity Measures in Networks

no code implementations25 Oct 2019 Palash Dey, Sourav Medya

These similarity measures turn out to be an important fundamental tool for many real world applications such as link prediction in networks, recommender systems etc.

Social and Information Networks Data Structures and Algorithms

Learning Heuristics over Large Graphs via Deep Reinforcement Learning

2 code implementations NeurIPS 2020 Sahil Manchanda, Akash Mittal, Anuj Dhawan, Sourav Medya, Sayan Ranu, Ambuj Singh

Additionally, a case-study on the practical combinatorial problem of Influence Maximization (IM) shows GCOMB is 150 times faster than the specialized IM algorithm IMM with similar quality.

Combinatorial Optimization Q-Learning +2

Cannot find the paper you are looking for? You can Submit a new open access paper.