1 code implementation • NeurIPS 2023 • Chanakya Ekbote, Ajinkya Pankaj Deshpande, Arun Iyer, Ramakrishna Bairi, Sundararajan Sellamanickam
This paper presents a simple filter-based augmentation method to capture different parts of the eigen-spectrum.
no code implementations • 23 Feb 2022 • Divyam Anshumaan, Sriram Balasubramanian, Shubham Tiwari, Nagarajan Natarajan, Sundararajan Sellamanickam, Venkata N. Padmanabhan
Simulating physical network paths (e. g., Internet) is a cornerstone research problem in the emerging sub-field of AI-for-networking.
1 code implementation • 7 Dec 2021 • Vijay Lingam, Chanakya Ekbote, Manan Sharma, Rahul Ragesh, Arun Iyer, Sundararajan Sellamanickam
We study various aspects of our proposed model including, dependency on the number of eigencomponents utilized, latent polynomial filters learned, and performance of the individual polynomials on the node classification task.
no code implementations • 29 Sep 2021 • Vijay Lingam, Chanakya Ajit Ekbote, Manan Sharma, Rahul Ragesh, Arun Iyer, Sundararajan Sellamanickam
We study various aspects of our proposed model including, dependency on the number of eigencomponents utilized, latent polynomial filters learned, and performance of the individual polynomials on the node classification task.
1 code implementation • ICLR 2022 • S Deepak Narayanan, Aditya Sinha, Prateek Jain, Purushottam Kar, Sundararajan Sellamanickam
Training multi-layer Graph Convolution Networks (GCN) using standard SGD techniques scales poorly as each descent step ends up updating node embeddings for a large portion of the graph.
no code implementations • 28 Jul 2021 • Vijay Lingam, Rahul Ragesh, Arun Iyer, Sundararajan Sellamanickam
More recent approaches adapt the eigenvalues of the graph.
no code implementations • 24 Jun 2021 • Vijay Lingam, Rahul Ragesh, Arun Iyer, Sundararajan Sellamanickam
Our approach achieves up to ~30% improvement in performance over state-of-the-art methods on heterophilic graphs.
Ranked #2 on Node Classification on Crocodile
no code implementations • 15 Feb 2021 • Rahul Ragesh, Sundararajan Sellamanickam, Vijay Lingam, Arun Iyer, Ramakrishna Bairi
CF-LGCN-U models naturally possess the inductive capability for new items, and we propose a simple solution to generalize for new users.
no code implementations • 19 Aug 2020 • Rahul Ragesh, Sundararajan Sellamanickam, Arun Iyer, Ram Bairi, Vijay Lingam
We consider the problem of learning efficient and inductive graph convolutional networks for text classification with a large number of examples and features.
no code implementations • 8 Apr 2020 • Rahul Ragesh, Sundararajan Sellamanickam, Vijay Lingam, Arun Iyer
Graph convolutional networks (GCNs) have gained popularity due to high performance achievable on several downstream tasks including node classification.
no code implementations • 1 Aug 2016 • Madhusudan Lakshmana, Sundararajan Sellamanickam, Shirish Shevade, Keerthi Selvaraj
Motivated by this observation, we propose to learn kernels with semantic coherence using clustering scheme combined with Word2Vec representation and domain knowledge such as SentiWordNet.
no code implementations • 10 Nov 2013 • Vinod Nair, Rahul Kidambi, Sundararajan Sellamanickam, S. Sathiya Keerthi, Johannes Gehrke, Vijay Narayanan
We consider the problem of quantitatively evaluating missing value imputation algorithms.
no code implementations • 9 Nov 2013 • Rahul Kidambi, Vinod Nair, Sundararajan Sellamanickam, S. Sathiya Keerthi
In this paper we propose a structured output approach for missing value imputation that also incorporates domain constraints.
no code implementations • 9 Nov 2013 • P. Balamurugan, Shirish Shevade, Sundararajan Sellamanickam
The optimization problem, which in general is not convex, contains the loss terms associated with the labelled and unlabelled examples along with the domain constraints.