Department of Computer Science and Engineering

Indian Institute of Technology Hyderabad (IITH) is a premier institute of science and technology established in 2008. IITH has been consistently ranked in the&nbsp;<a href="https://iith.ac.in/reports/" rel="noopener noreferrer">top 10 institutes</a>&nbsp;in India for Engineering&nbsp;<a href="https://www.nirfindia.org/2021/EngineeringRanking.html" rel="noopener noreferrer" target="_blank">according to NIRF</a>&nbsp;making it one of the most coveted schools for science and technology in the country.

IITH offers&nbsp;<a href="https://iith.ac.in/academics/under-graduate/">undergraduate programs</a>&nbsp;in all the classical engineering disciplines, applied sciences, design, as well as several modern interdisciplinary areas. Students are given a flexibility to explore a broad set of areas, and potentially pursue a minor or double major in a discipline that is not their own. Students who wish to seek a deeper understanding of their own discipline are strongly encouraged to get involved in cutting-edge research with the help of a faculty to mentor them, and earn an Honors in their own field.

The very foundation of IIT Hyderabad is based on&nbsp;<a href="https://iith.ac.in/research/">research and innovation</a>. The vibrant research culture is evident from the number of patents and publications that IITH produces consistently every year. IITH offers&nbsp;<a href="https://iith.ac.in/academics/programmes-offered/">graduate programs</a>&nbsp;at both a masters, and a doctoral level, in several diverse areas. There are separate programs for technology, design, science, and liberal arts. The&nbsp;<a href="https://iith.ac.in/academics/post-graduate/">MTech program</a>&nbsp;is offered in various modes and durations to cater to the ever-growing need of postgraduate level professionals in the country.

IITH encourages and supports&nbsp;<a href="https://iic-iith.netlify.app/index.html" rel="noopener noreferrer" target="_blank">innovation and entrepreneurship</a>&nbsp;at every stage. The minor program in entrepreneurship is open to all students of IITH, and many of the courses are offered by industrialists who are entrepreneurs themselves. There are various&nbsp;<a href="https://iith.ac.in/research/centres-incubators/">innovation centers and incubators</a>&nbsp;that promote entrepreneurship and help mentor young innovators. IITH has been very successful in building tie-ups with leading academic institutions around the globe.

IITH enjoys a very special relationship with Japanese universities and industries that goes beyond academic and research collaborations. In fact, some of the iconic buildings in IITH campus will carry the signature of Japanese architecture.

IITH creates a unique holistic ecosystem for education that offers interactive learning, a very flexible academic structure, cutting-edge research, strong industry collaboration, and entrepreneurship. This is an environment which enables students and faculty to translate their dreams into realities.

IIT Hyderabad

Momentum based learning algorithms are one of the most successful learning algorithms in both convex and non-convex optimization. Two major momentum based techniques that achieved tremendous success in gradient-based optimization are Polyak’s heavy ball method and Nesterov’s accelerated gradient. A crucial step in all the momentum based methods is the choice of the momentum parameter m, which is always set to less than 1. Although the choice of m < 1 is justified only under very strong theoretical assumptions, it works well in practice. In this paper we propose a new momentum based method ADINE, which relaxes the constraint of m < 1 and allows the learning algorithm to use adaptive higher momentum. We motivate our relaxation on m by experimentally verifying that a higher momentum (? 1) can help escape saddles much faster. ADINE uses this intuition and helps weigh the previous updates more, inherently setting the momentum parameter to be greater in the optimization method. To the best of our knowledge, the idea of increased momentum is first of its kind and is very novel. We evaluate this on deep neural networks and show that ADINE helps the learning algorithm to converge much faster without compromising on the generalization error. © 2018 Association for Computing Machinery.

World University root node

ACM International Conference Proceeding Series

ADINE: An adaptive momentum method for stochastic gradient descent

This paper proposes two network architectures to perform video generation from captions using Variational Autoencoders. We adopt a new perspective towards video generation where we use attention as well as allow the captions to be combined with the long-term and short-term dependencies between video frames and thus generate a video in an incremental manner. Our experiments demonstrate our network architectures’ ability to distinguish between objects, actions and interactions in a video and combine them to generate videos for unseen captions. Our second network also exhibits the capability to perform spatio-temporal style transfer when asked to generate videos for a sequence of captions. We also show that the network’s ability to learn a latent representation allows it generate videos in an unsupervised manner and perform other tasks such as action recognition. © 2018, Springer Nature Singapore Pte Ltd.

Communications in Computer and Information Science

From Recognition to Generation Using Deep Learning: A Case Study with Video Generation

The vertical interplay among Edge-Fog-Cloud computing introduces a new requirement to accommodate services, network functions and chains of them in a heterogeneous environment. However, such operation involves challenging tasks including management of resource availability, compliance with Service Level Agreements (SLAs), dynamic re-configuration of services and networks across different computing domains. This paper proposes Dynamic Network Function (DNF) that involves 1) a tractable VNF placement algorithm, and 2) a bottleneck detection and removal algorithm for the heterogeneous computing and network environments representing the Edge-Fog-Cloud interplay. The proof of concept (PoC) implementation of DNF has been done using Docker-based NFV and SDN setups as well as a simulation setup for studying the DNF scalability. The evaluation results showed that DNF could achieve high acceptance ratio of SLA requests with reasonably short duration. The PoC implementation also exhibited the proper functions to 1) orchestrate the NFV and SDN, and 2) detect and mitigate the bottlenecks. © 2018 Association for Computing Machinery.

Asian Internet Engineering Conference, AINTEC 2018

Dynamic Network Function (DNF): Service chain placement and bottleneck removal for edge-fog-cloud interplay

With the rise of user-generated content in social media coupled with almost non-existent moderation in many such systems, aggressive contents have been observed to rise in such forums. In this paper, we work on the problem of aggression detection in social media. Aggression can sometimes be expressed directly or overtly or it can be hidden or covert in the text. On the other hand, most of the content in social media is non-aggressive in nature. We propose an ensemble based system to classify an input post into one of three classes, namely, Overtly Aggressive, Covertly Aggressive, and Non-aggressive. Our approach uses three deep learning methods, namely, Convolutional Neural Networks (CNN) with five layers (input, convolution, pooling, hidden, and output), Long Short Term Memory networks (LSTM), and Bi-directional Long Short Term Memory networks (Bi-LSTM). A majority voting based ensemble method is used to combine these classifiers (CNN, LSTM, and Bi-LSTM). We trained our method on Facebook comments dataset and tested on Facebook comments (in-domain) and other social media posts (cross-domain). Our system achieves the F1-score (weighted) of 0.604 for Facebook posts and 0.508 for social media posts. © COLING 2018 - 1st Workshop on Trolling, Aggression and Cyberbullying, TRAC 2018 - Proceedings of the Workshop.

COLING 2018 - 1st Workshop on Trolling, Aggression and Cyberbullying, TRAC 2018 - Proceedings of the Workshop

Aggression Detection in Social Media using Deep Neural Networks

Location Based Social Networks (LBSN) benefit the users by allowing them to share their locations and life moments with their friends. The users can also review the locations they have visited. Classical recommender systems provide users a ranked list of single items. This is not suitable for applications like trip planning,where the recommendations should contain multiple items in an appropriate sequence. The problem of generating such recommendations is challenging due to various critical aspects, which includes user interest, budget constraints and high sparsity in the available data used to solve the problem. In this paper, we propose a graph based approach to recommend a set of personalized travel packages. Each recommended package comprises of a sequence of multiple Point of Interests (POIs). Given the current location and spatio-temporal constraints, our goal is to recommend a package which satisfies the constraints. This approach utilizes the data collected from LBSNs to learn user preferences and also models the location popularity. © 2018 Association for Computing Machinery.

UMAP 2018 - Adjunct Publication of the 26th Conference on User Modeling, Adaptation and Personalization

Personalized tourist package recommendation using graph based approach

Given an ideal a in A[x1,…,xn] where A is a Noetherian integral domain, we propose an approach to compute the Krull dimension of A[x1,…,xn]/a, when the residue class ring is a free A-module. When A is a field, the Krull dimension of A[x1,…,xn]/a has several equivalent algorithmic definitions by which it can be computed. But this is not true in the case of arbitrary Noetherian rings. For a Noetherian integral domain A we introduce the notion of combinatorial dimension of A[x1,…,xn]/a and give a Gröbner basis method to compute it for residue class rings that have a free A-module representation w.r.t. a lexicographic ordering. For such A-algebras, we derive a relation between Krull dimension and combinatorial dimension of A[x1,…,xn]/a. An immediate application of this relation is that it gives a uniform method, the first of its kind, to compute the dimension of A[x1,…,xn]/a without having to consider individual properties of the ideal. For A-algebras that have a free A-module representation w.r.t. degree compatible monomial orderings, we introduce the concepts of Hilbert function, Hilbert series and Hilbert polynomials and show that Gröbner basis methods can be used to compute these quantities. We then proceed to show that the combinatorial dimension of such A-algebras is equal to the degree of the Hilbert polynomial. This enables us to extend the relation between Krull dimension and combinatorial dimension to A-algebras with a free A-module representation w.r.t. a degree compatible ordering as well. © 2017 Elsevier Ltd

Journal of Symbolic Computation

On Gröbner bases and Krull dimension of residue class rings of polynomial rings over integral domains

Welcome Message from the General and PC Chairs of ICVGIP 2018

Reduction in transistor size has led to increase in transistor density allowing multiple cores to be placed on the same die; thus increasing the power dissipation of the chip. The heat dissipation by the cores may cause the chip temperatures to rise beyond a threshold limit causing it to malfunction. Dynamic Thermal Management (DTM) techniques address this issue by checking the rising temperature of the cores in the CMPs. Current techniques available for DTM use either Dynamic Frequency Scaling (DFS) or TM. The DFS based technique controls the chip temperature by reducing the cores operating frequency; however this may affect its performance. The technique of task rotation incurs the overheads of application migration. In this paper we propose a DTM technique that combines the benefits of both DFS and TM. The aim of the approach is to facilitate the efficient execution of high priority tasks. For this we allow other low priority tasks to sacrifice their performance to control the overall chip temperature. In particular, the high priority task is made to run at maximum allowable speed and, if needed, migrated to less heated cores while maintaining its performance. In this process, a low priority task runs at lower frequency to aid temperature reduction. Experimental analysis found that our approach reduces peak temperature of chip by 15.62 degree C with 0.24 percent performance (MIPS) degradation of high priority tasks compared to respective baseline. © 2018 IEEE.

Proceedings of the IEEE International Conference on VLSI Design

Dynamic thermal management by using task migration in conjunction with frequency scaling for chip multiprocessors

Hawkes Processes are probabilistic models useful for modelling the occurrences of events over time. They exhibit mutual excitation property, where a past event influences future events. This has been successful in modelling the evolution of memes and user behaviour in social networks. In the Hawkes process, the occurrences of events are determined by an underlying intensity function which considers the influence from past events. The intensity function models the mutual-exciting nature by adding up the influence from past events. The calculation of the intensity function for every new event requires time proportional to the number of past events. When the number of events is high, the repeated intensity function calculation will become expensive. We develop a faster approach which takes only constant time complexity to calculate the intensity function for every new event in a mutually exciting Hawkes process. This is achieved by developing a recursive formulation for mutually exciting Hawkes process and maintaining an additional data structure which takes a constant space. We found considerable improvement in runtime performance of the Hawkes process applied to the sequential stance classification task on synthetic and real world datasets. © 2018 IEEE.

2018 10th International Conference on Communication Systems and Networks, COMSNETS 2018

Accelerating Hawkes process for event history data: Application to social networks and recommendation systems

In order to sustain the user-base for a web service, it is important to know the return time of a user to the service. We propose a Bayesian point process, log Gaussian Cox process (LGCP), to model and predict return time of users. It allows encoding the prior domain knowledge and non-parametric estimation of latent intensity functions capturing user behaviour. We capture the similarities among the users in their return time by using a multi-task learning approach. We show the effectiveness of the proposed approaches on predicting the return time of users to last.fm music service. © 2018 Association for Computing Machinery.