Department of Data Science

Indian Institute of Technology Palakkad&nbsp;(IIT Palakkad or IIT PKD) is a public autonomous&nbsp;engineering&nbsp;and&nbsp;research&nbsp;institute located in&nbsp;Palakkad,&nbsp;Kerala. It was&nbsp;one of the five new&nbsp;IITs&nbsp;proposed in the&nbsp;2014 Union budget of India. The campus was inaugurated on 3 August 2015. The institute has 94 faculty, 978 students, and 63 non-teaching staffs.

IT Palakkad currently offers four-year Bachelor of Technology (B.Tech) and two years MS programmes in 8 different engineering disciplines including&nbsp;Computer Science, Electrical, Mechanical, Civil among others.

&nbsp;

IIT Palakkad

Studies of networked phenomena, such as interactions in online social media, often rely on incomplete data, either because these phenomena are partially observed, or because the data is too large or expensive to acquire all at once. Analysis of incomplete data leads to skewed or misleading results. In this paper, we investigate limitations of learning to complete partially observed networks via node querying. Concretely, we study the following problem: given (i) a partially observed network, (ii) the ability to query nodes for their connections (e.g., by accessing an API), and (iii) a budget on the number of such queries, sequentially learn which nodes to query in order to maximally increase observability. We call this querying process Network Online Learning and present a family of algorithms called NOL*. These algorithms learn to choose which partially observed node to query next based on a parameterized model that is trained online through a process of exploration and exploitation. Extensive experiments on both synthetic and real world networks show that (i) it is possible to sequentially learn to choose which nodes are best to query in a network and (ii) some macroscopic properties of networks, such as the degree distribution and modular structure, impact the potential for learning and the optimal amount of random exploration. © 2020, The Author(s).

World University root node

Fulltext

Applied Network Science

Understanding the limitations of network online learning

In this work, we consider a sequence of J matrix multiplication jobs which needs to be distributed by a master across multiple worker nodes. For i ? {1, 2, . . ., J}, job-i begins in round-i and has to be completed by round-(i + T). Previous works consider only the special case of T = 0 and focus on coding across workers. We propose here two schemes with T > 0, which feature coding across workers as well as the dimension of time. Our first scheme is a modification of the polynomial coding scheme introduced by Yu et al. and places no assumptions on the straggler model. Exploitation of the temporal dimension helps the scheme handle a larger set of straggler patterns than the polynomial coding scheme, for a given computational load per worker per round. The second scheme assumes a particular straggler model to further improve performance (in terms of encoding/decoding complexity). We develop theoretical results establishing (i) optimality of our proposed schemes for a certain class of straggler patterns and (ii) improved performance for the case of i.i.d. stragglers. These are further validated by experiments, where we implement our schemes to train neural networks. © 2020 Neural information processing systems foundation. All rights reserved.

Advances in Neural Information Processing Systems

Coded sequential matrix multiplication for straggler mitigation

Recently generative models have focused on combining the advantages of variational autoencoders (VAE) and generative adversarial networks (GAN) for good reconstruction and generative abilities. In this work we introduce a novel hybrid architecture, Implicit Discriminator in Variational Autoencoder (IDVAE), that combines a VAE and a GAN, which does not need an explicit discriminator network. The fundamental premise of the IDVAE architecture is that the encoder of a VAE and the discriminator of a GAN utilize common features and therefore can be trained as a shared network, while the decoder of the VAE and the generator of the GAN can be combined to learn a single network. This results in a simple two-tier architecture that has the properties of both a VAE and a GAN. The qualitative and quantitative experiments on real-world benchmark datasets demonstrate that IDVAE performs better than the state of the art hybrid approaches. We experimentally validate that IDVAE can be easily extended to work in a conditional setting and demonstrate its performance on complex datasets.

Proceedings of the International Joint Conference on Neural Networks

Implicit Discriminator in Variational Autoencoder

In this paper, we design erasure-correcting codes for channels with burst and random erasures, when a strict decoding delay constraint is in place. We consider the sliding-window-based packet erasure model proposed by Badr et al., where any time-window of width w contains either up to a random erasures or an erasure burst of length at most b. One needs to recover any erased packet with a strict decoding delay deadline of tau , where erasures are as per the channel model. Presently existing rate-optimal constructions in the literature require, in general, a field-size which grows exponential in tau , as long as frac {a}{tau } remains a constant. In this work, we present a new rate-optimal code construction covering all channel and delay parameters, which requires an O(tau {2}) field-size. As a special case, when (b-a)=1 , we have a field-size linear in tau . We also present two other constructions having linear field-size, under certain constraints on channel and decoding delay parameters. As a corollary, we obtain low field-size, rate-optimal convolutional codes for any given column distance and column span. Simulations indicate that the newly proposed streaming code constructions offer lower packet-loss probabilities compared to existing schemes, for selected instances of Gilbert-Elliott and Fritchman channels. © 1963-2012 IEEE.

IEEE Transactions on Information Theory

Rate-Optimal Streaming Codes for Channels with Burst and Random Erasures

In this paper, we propose a novel CGRA architecture and associated compilation flow supporting both integer and floating-point computations for energy efficient acceleration of DSP applications. Experimental results show that the proposed accelerator achieves a maximum of 4.61× speedup compared to a DSP optimized, ultra low power RISC-V based CPU while executing seizure detection, a representative of wide range of EEG signal processing applications with an area overhead of 1.9 × The proposed CGRA achieves a maximum of 6.5× energy efficiency compared to the CPU. © 2020 Institute of Electrical and Electronics Engineers Inc.. All rights reserved.

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

Energy efficient acceleration of floating point applications onto CGRA

In this paper, we present a novel unsupervised domain adaptation framework, Multi-Partition Feature Alignment Network, that learns a deep neural model for the target domain without the need for any supervision. Recent leading approaches for unsupervised domain adaptation are based on adversarial alignment. Aligning the global distribution of the domain representations via adversarial training does not guarantee the class-wise distribution alignment. The proposed approach is built on adversarial learning with the focus on carefully aligning class-wise domain representations. Our algorithm utilizes the pseudo-labels (the predicted labels) of the target features to stimulate class-wise alignment. As the pseudo-labels of individual target features can be erroneous, instead of iteratively aligning individual target samples, the proposed framework introduces a generic class-specific multi-partition alignment procedure that enables superior class-discriminative alignment of domain representations. The competitive performance of the proposed framework against state-of-the-art approaches over a wide variety of visual recognition tasks, namely, the digits classification task and the object recognition task, validates its effectiveness for unsupervised domain adaptation.

Multi-Partition Feature Alignment Network for Unsupervised Domain Adaptation

Similar facts are often expressed in different ways in natural language text, which introduces the redundancy and ambiguity of Subject-Predicate-Object (SPO) triples in Open Information Extraction (Open IE). This work focuses on canonicalizing such SPO triples. We propose a clustering framework using non-negative matrix tri-factorization that jointly clusters predicate phrases and subject-object pairs and aligns them in a meaningful manner. The evaluation shows that our co-clustering method outperforms significantly over rule mining and Knowledge-Base-embedding approaches for two existing datasets.

PervasiveHealth: Pervasive Computing Technologies for Healthcare

Co-clustering triples from open information extraction

Quantities are more than numeric values. They represent measures for entities, expressed in numbers with associated units. Search queries often include quantities, such as athletes who ran 200m under 20 seconds or companies with quarterly revenue above $2 Billion. Processing such queries requires understanding the quantities, where capturing the surrounding context is an essential part of it. Although modern search engines or QA systems handle entity-centric queries well, they consider numbers and units as simple keywords and therefore fail to understand the condition (less than, above, etc.), the unit of interest (seconds, dollar, etc.) and the context of the quantity (200m race, quarterly revenue, etc.) As a result, they cannot generate the correct candidate answers. In this work, we demonstrate a prototype QA system, called Qsearch, that can handle advanced queries with quantity constraints using the common cues present in both query and the text sources.

WSDM 2020 - Proceedings of the 13th International Conference on Web Search and Data Mining

Entities with quantities: Extraction, search and ranking

Locally recoverable codes (LRCs) are known to reduce the repair locality, i.e., the number of nodes accessed during node repair, in distributed storage systems. Regenerating codes, on the other hand, offer a decreased repair bandwidth, which is the repair traffic incurred during node repairs. Locally regenerating codes (LRGCs), which are constructed via intertwining regenerating codes of smaller block lengths, simultaneously offer a small repair locality and repair bandwidth. In this letter, we construct a family of LRGCs where the constituent codes are bandwidth-efficient, minimum bandwidth cooperative regenerating codes. The newly constructed LRGCs are optimal with respect to both minimum distance and rate. © 1997-2012 IEEE.

IEEE Communications Letters

Codes with minimum bandwidth cooperative local regeneration

An (a, b, τ) streaming code is a packet-level erasure code that can recover under a strict delay constraint of τ time units, from either a burst of b erasures or else of a random erasures, occurring within a sliding window of time duration w. While rate-optimal constructions of such streaming codes are available for all parameters {a, b, τ, w} in the literature, they require in most instances, a quadratic, O(τ2) field size. In this work, we make further progress towards field size reduction and present rate-optimal O(τ) field size streaming codes for two regimes: (i) gcd(b, τ + 1 - a) ≥ a (ii) τ + 1 a + b and b mod a 0, a - 1. © 2020 IEEE.