Abstract

Today human identification is a practice followed quite frequently in contexts ranging from forensic labs to high security organisations to almost every domestic household. One of the oldest and most accurate ways to establish the identity of an individual is through fingerprints. It is so because, biologically no two people (not even identical twins) have the same fingerprints. Hence it comes as no surprise that fingerprint biometrics has been a very active domain of research for a long time. Over the years, especially with the emergence of the deep learning era, the computational modules of a fingerprint recognition system have become bigger and bigger. Simultaneously, the actual hardware deploying fingerprint recognition systems has become more and more user friendly. These developments provide a contrast as one tries to incorporate the highly accurate but also extremely heavy modules of a fingerprint recognition system in the palm of our hands. In this thesis, we will perform a couple of studies which explore ways to optimize fingerprint recognition systems. This is done by trying to parallelize modules and eliminate semantically redundant computations. Along the way we will also obtain better performances as compared to the academic state of the art. Typical fingerprint recognition systems are comprised of a spoof detection module and a subsequent matching module, running one after the other. In the first study, we posit that both spoof detection and fingerprint matching are correlated tasks. Therefore, rather than performing the two tasks separately, we propose a joint model for spoof detection and matching to simultaneously perform both tasks without compromising the accuracy of either task. In practice, this reduces the time and memory requirements of the fingerprint recognition system by 50% and 40%, respectively. On the other hand, fingerprint matching task itself can be solved in multiple ways. Fingerprint feature extraction is a core step in the fingerprint matching task and it can be solved by using a global as well as a local approach. State of the art global approaches use huge deep learning models to process the full fingerprint image at once which makes the corresponding approach memory intensive. Whereas local approaches involve minutiae based patch extraction and multiple feature extraction steps which make the corresponding approach time-intensive. But both these approaches provide useful and sometimes exclusive insights for solving the problem. Hence using both kinds of approaches together for extracting fingerprint representations is semantically useful but quite inefficient. Hence, in the second study, we use a convolutional transformer based approach with an in-built minutiae extractor. This study provides an efficient solution for feature extraction that parallely provides a global as well as a local represen tation of a fingerprint. This not only provides a replacement for using a heavy combination of two independent local and global approaches but also provides a further 54.41% speedup over the local approach and a 57.93% conservation in memory as compared to the global approach. Both of these studies work towards the common goal of optimizing fingerprint recognition systems and hence provide a significant advantage in the context of such systems deployed on resource constrained devices like smartphones.

Abstract

Recent work in NLP has focused on applying powerful neural sequence models to various learning problems. These neural models excel at extracting statistical patterns from large amounts of data but struggle to learn patterns or reason about language from only a few examples. We ask the question: Can we learn explicit rules that generalize well from only a few examples? We explore this question by viewing linguistic rules as programs that operate on linguistic forms. This allows us to tackle the problem of learning linguistic rules using program synthesis – a method to learn rules in the form of programs in a domain-specific language (DSL). We develop a synthesis model to learn phonological rules as programs in a DSL. In addition to being highly sample-efficient, our approach generates human-readable programs and allows control over the generalizability of the learned programs. We test the ability of our models to generalise from only a few training examples using our new dataset of problems from the Linguistics Olympiad. These problems are tasks from contests for high school students around the world that require inferring linguistic patterns from a small number of given examples. These problems are a challenging set of tasks that require strong linguistic reasoning ability. Having shown that program synthesis can be used to learn phonological rules in highly data-constrained settings, we use the problem of phonological stress placement as a case to study how the design of the domain-specific language influences the generalisation ability when using the same learning algorithm. We find that encoding the distinction between consonants and vowels results in much better performance, and providing syllable-level information further improves generalization. Program synthesis, thus, provides a way to investigate how access to explicit linguistic information influences what can be learned from a small number of examples.

Abstract

Virtual Labs are a collection of college-level online experiments in the science and engineering disciplines. Virtual Labs provides a remote experimentation platform for hands-on learning. It is essential to analyze and improve the web performance of the Virtual Labs platform to enhance the user experience as well as to cope with the increased usage. This study identifies and implements approaches to improve the performance of Virtual Labs experiments. We gather the web performance data using two tools, Google Analytics and Webpagetest, analyze and implement the performance improvement approaches. The implemented approaches have reduced the load time of the Virtual Labs experiments by 80% and reduced the load time of the page’s main content from 6.1 seconds to under 1.5 seconds. The loading and rendering metrics have been optimized, and the efficiency of the experiment loading has increased.

Abstract

Music, an integral part of our lives, which is not only a source of entertainment but plays an important role in mental well-being by impacting moods, emotions and other affective states. In this age of Big Data, online music streaming services allow us to capture ecologically valid music listening behavior and provide a rich source of information to identify several user-specific aspects. Music preferences and listening strategies have been shown to be associated with the psychological well-being of listeners including internalized symptomatology and depression. However, till date no studies exist that examine music-related social tags, static & time-varying audio-derived features, in terms of acoustic and emotional content, and its association with users well-being. The current work aims at unearthing static and dynamic patterns and trends in the individuals at risk for depression as it manifests in naturally occurring online music consumption. Mental well-being scores and listening histories of 541 Last.fm users were examined. Social tags associated with each listener’s most popular tracks were analyzed to unearth the mood/emotions and genres associated with the users. Results revealed that social tags prevalent in the users at risk for depression were predominantly related to emotions depicting Sadness associated with genre tags representing neo-psychedelic-, avant garde-, dream-pop. Static and dynamic acoustic and emotion-related features were extracted from each user’s listening history and correlated with their mental well-being scores. Results revealed that individuals with greater depression risk resort to higher dependency on music with greater repetitiveness in their listening activity. Furthermore, the affinity of depressed individuals towards music that can be perceived as sad was found to be resistant to change over time. This study has large implications for future work in the area of assessing mental illness risk by exploiting digital footprints of users via online music streaming platforms. This study will open up avenues for an MIR-based approach to characterizing and predicting risk for depression which can be helpful in early detection and additionally provide bases for designing music recommendations accordingly

Abstract

Popularity prediction is a well-studied machine learning task with wide-ranging business applications. The primary goal in a popularity prediction problem is to estimate the future popularity of a piece of content that is put up on the Internet. For example, a popularity prediction model might estimate the number of page views that a particular online news article would accrue over a time period. Or a social-media popularity prediction system might try to foretell the number of comments or shares that some social media post would receive over a week. Being able to predict the prospective popularity of content within a very short time after its publication has multiple downstream applications in article recommendation and advertising, and document retrieval. Additionally, proactively forecasting document popularity effectively allows an editor to improve a document in order to make it more engaging. Though there has been much work on the document-level prediction of prospective popularity labels for news articles, online petitions, social media posts, etc., there have been no attempts to forecast information popularity at a more granular level. Sentence-level popularity forecasting is of interest not only because of the insights it produces about how informative sentences contribute to a document’s overall popularity but also because of its potential applications in tasks such as popularity-guided text summarization, pull quote selection, catchphrase extraction, webpage search engine optimization, etc. In this thesis, we overcome existing challenges to capture information popularity at the level of individual sentences, and we present neural models that effectively forecast sentence-specific information popularity. Formally, we introduce the task of proactively forecasting relative popularities of sentences within online news documents solely utilizing their natural language content. Instead of taking a simplistic route where one would predict binary popular-or-not labels for individual sentences, we frame sentencepopularity forecasting as a regression task where we assign a normalized score ∈ [0, 1] to each sentence indicating its intra-document information popularity relative to all other sentences. To overcome the absence of sentence-level labeled data for information popularity, we curate INFOPOP, a dataset containing popularity labels for over 1.7 million sentences from over 50,000 online news documents sourced from 26 reputed news websites. To the best of our knowledge, INFOPOP is the first dataset automatically created using streams of incoming search engine queries to generate sentence-level popularity annotations. We observe how certain aspects of information popularity are related to sentence salience and propose auxiliary transfer learning subtasks of sentence salience prediction for pretraining our neural sentence popularity forecasting models. Sentence salience is well studied in the context of automatic text summarization. In this thesis, we provide an overview of document summarization approaches and present our findings from two scholarly document summarization tasks. In particular, we describe an abstractive and an extractive approach for summarizing scientific research papers, which have led to state-of-the-art results on two shared tasks, namely LaySumm and LongSumm, at the Scholarly Document Processing workshop (SDP) at EMNLP ’20. We further discuss the distinction between salience and sentences’ summary-inclusion worthiness in an extractive summarization setting. For sentence-specific popularity forecasting, we experiment with unsupervised baselines, a rudimentary recurrent neural network-based model, and a robust sentence-sequence regression architecture based on BERT (Bidirectional Encoder Representations from Transformers). Our results show that our proposed Transfer Learning approach produces significant performance gains on all metrics – reaching nDCG values over 0.8. We conduct extensive evaluations to highlight insightful distinctions between our primary and auxiliary tasks. Notably, our findings showcase a non-trivial takeaway: though information popularity and information salience are varying concepts, transfer learning from salience prediction significantly enhances sentence popularity forecasting.

Abstract

Dementia is a neurological disease that affects people starting around the age of 45. The most common symptom of dementia includes memory difficulties, impaired judgment-making capacity, difficulty in performing daily living activities, mind wandering, and disorientation (WHO 2022). Researchers have been looking into the causes, rate of progression, and potential remedies for many years. However, there is as yet no cure for dementia. As a result, researchers have been investigating the efficacy of alternative therapeutic methods such as regular exercise, therapies, meditation, etc. The present thesis has focused on investigating dementia Alzheimer's type and its early phases. It includes an overview of Alzheimer's disease and potential research gaps related to atrophy rates in selected regions of interest, such as the hippocampal subfields and the effects of meditation as adjuvant therapy. We used longitudinal structural and functional MRI datasets from humans to address these two research topics in the thesis. We have explored the nature of atrophy in the hippocampal subfields during the distinct phases of Alzheimer's disease. This research hypothesized that hippocampal subfields in Alzheimer's disease patients would have the highest gray matter volume atrophy rate compared to other groups (control, early MCI, and late MCI). We have also proposed that the subfields being involved in memory encoding and decoding would deteriorate faster in the late MCI or AD groups. The analysis used structural MRI data available at two-time points (separated by a year) from a cohort of 196 participants in the publicly available ADNI dataset. We have estimated the gray matter volume (as the cortical thickness is irrelevant here in the subcortical region) and compared it across the four groups. Subfields that are mainly involved in the process of memory encoding, consolidation, and decoding, such as the dentate gyrus, CA3, CA1, and subiculum, have shown a higher atrophy rate in Alzheimer's disease and late MCI groups as compared to early MCI or healthy controls over a one-year time period.We have also investigated the long-term effects of meditation techniques (one of the alternative therapy approaches) in early MCI or Alzheimer's disease patients over the course of six months. This study hypothesized that simple long-term daily meditation would have a facilitatory effect on the brain regions related to attention and executive control in patients with early MCI or Alzheimer's disease. In his study, we have used a dataset collected by our collaborators on a cohort of dementia patients from Kolkata, India. In this longitudinal study, we compared the meditation and non-meditation groups on two structural brain measures: cortical thickness (CT) and gray matter volume (GMV). Our results indicate that long-term daily meditation led to substantial improvements in cortical thickness and gray matter volume in the regions responsible for attention and executive control in the prefrontal and inferior parietal cortex. In summary, the thesis presents results from two longitudinal studies. One study characterized the atrophy rate differences among aged controls and early-to-late-stage dementia patients. The other study investigated the efficacy of meditation in arresting the structural decline in some critical brain regions. It is hoped that these studies are useful for the clinical management of dementia.

Abstract

Machine learning techniques have found widespread usage in critical systems in the recent past.This rise of ML was mainly fueled by Neural Networks, which enabled end-to-end systems to use.This is possible because of the feature learning and universal approximator properties of NNs. Recentresearch has extended the application of NNs to the estimation of mutual information (MI) from samples.They construct a critic using a NN and maximize a variation lower bound of MI. These methods havealleviated many problems faced by previous MI estimation methods like the inability to scale andincompatibility with mini-batch-based optimization. While these estimation methods are reliable whenthe true MI is low, they tend to produce high variance estimates when the true MI is high. We arguethat the high variance characteristics are due to the uncontrolled complexity of the critic’s hypothesisspace. In support of this argument, we use the data-drivenRademacher complexityof the hypothesisspace associated with the critic’s architecture to derive upper bounds on thegeneralization errorofvariational lower bound based estimates of MI. Using the derived bounds, we show that Rademachercomplexity of the critic’s hypothesis is a major contributor to the generalization error. Inspired by thiswe propose to negate the high variance characteristics of these estimators by constraining the critic’shypothesis space to aReproducing Kernel Hilbert Space(RKHS), which corresponds to a kernel learnedusingAutomated Spectral Kernel Learning(ASKL). Further, we propose two regularisation terms byexamining the upper bound on the Rademacher complexity of RKHS learned by ASKL. The weightsof these regularization terms can be varied for a bias-variance tradeoff. We empirically demonstratethe efficacy of this regularization in enforcing proper bias-variance tradeoff on four different variationallower bounds of MI, namely NWJ, MINE, JS and SMILE.We also propose a novel application of MI estimation using variational lower bounds and NNs inlearning disentangled representations of videos. The learned representations contain two sets of features,content and pose representations which disjointly represent the content and movement in the video. Thisis achieved by minimizing the mutual information between pose representation of different frames from the same video, thus in effect forcing the content representation to capture information common betweendifferent frames. We qualitatively and quantitatively compare our method against DRNET [24] on twosynthetic and one real-world dataset.

Abstract

As we all know, training a highly effective online learning model for complex tasks often hinges on the abundance of high-quality noise-free labeled data. However, acquiring this high-quality, noise-free labeled data is becoming a bottleneck in cost, time, and computational resources. In this thesis, we make a sincere effort to address two significant issues that the current State-Of-TheArt bandit feedback-based online learning algorithms fail to address: (a) noise present in bandit feedback and (b) algorithms’ heavy reliance on labeled data. To deal with the noise present in the bandit feedback, we proposed a novel algorithm named RCINE, robust to noisy bandit feedback. The methodology used in RCINE requires knowledge of the noise rates. We proposed a subroutine called NREst to estimate the noise rates, resulting in an end-to-end learning framework for learning a multiclass classifier under noisy bandit feedback. The proposed algorithm enjoys a mistake bound of the order of O( √ T) in the high noise case and of the order of O(T 2/3 ) in the worst case. We also show our approach’s effectiveness using extensive experiments on several benchmark datasets. Furthermore, to reduce the reliance of an online supervised learning algorithm on labeled data, we proposed ALBIF, an efficient stochastic sub-gradient descent algorithm for learning a multiclass classifier under an active bandit feedback setting. ALBIF enjoys a regret bound of the order O(log T) in the active learning setting as well as in the standard (non-active) bandit feedback setting. We also demonstrate the effectiveness of the proposed algorithms by conducting extensive experiments on various real-world and synthetic datasets against several benchmark algorithms.

Abstract

The rise in energy demand is a major cause of global warming and climate change. The U.S. Energy Information Administration (EIA) projects that the world energy consumption will grow by nearly 50% and the energy consumed in the building sector will increase by around 65% by 2050. There are two crucial sustainability measures, improving energy efficiency and reducing energy consumption in the building sector. To assist in this, Appliance Load Monitoring (ALM) is essential for effective load management because it determines the energy consumption and operating states of individual appliances. Energy feedback information can help reduce consumption by 5% to 15%. Compared to a single aggregate consumption feedback, a detailed individual appliance consumption feedback is more beneficial. Most electrical appliances do not have the capabilities to monitor their individual energy consumption details. To estimate this energy consumption for appliances, the two most widely used load monitoring approaches are Intrusive Load Monitoring (ILM) and Non-Intrusive Load Monitoring (NILM). ILM makes use of distributed sensing using smart power strips or power sockets to monitor individual appliances. Such kind of installation requires entering inside a house and therefore, ILM is intrusive in nature. NILM or load disaggregation is an approach to estimate individual appliance energy consumption from aggregate load measurement obtained from a single measurement point. Features extracted from electrical signals are used for automatic appliance identification. Researchers in the past have published machine learning-based ILM and NILM techniques. However, these machine learningbased ILM/NILM techniques are sensitive to noise, and their performance degrades in the presence of variations in data due to noise, sensor drift and source voltage fluctuations. The observations made by ESMI (Electrical Supply Monitoring Initiate) of Prayas Group show that grid voltages fluctuate approximately between 210 to 250 V in some regions of India. Varying voltages lead to changes in appliance power consumption patterns. Appliance identification models may not recognize appliances correctly because of their modified consumption patterns. The ILM and NILM approaches have to be tested in these varying operating conditions. There is a need to develop a low-cost, efficient and accurate load monitoring solution. This thesis investigates the robustness of supervised load identification techniques for improving ILM and NILM approaches. The first part of the thesis provides techniques to improve the ILM approach, especially to monitor plug loads. Plug loads are the appliances that are plugged in AC sockets. Plug loads account for 20% to 30% of building energy consumption, and there is an increasing trend in plug load consumption. The energy performance of buildings can be improved by effective plug load monitoring and control. A plug load monitoring field study is conducted in an academic institute to gain insights into plug load usage and energy consumption patterns. Large scale deployment of such load monitoring solutions is costly. Without load-sensing capabilities, it isn’t easy to track some plug loads that keep changing locations. Smart plug strips or smart sockets with plugged-in load identification capability are helpful for monitoring loads automatically. Smart strips or sockets record electrical measurements using an energy metering sensor. The extracted features or load signatures are used to train machine learning models for load identification. Comparative performance analysis of these load identification techniques is provided. The results show that a simple algorithm like KNN on low-frequency data of two minutes provides excellent identification accuracy. However, the identification models had problems in the presence of voltage variations. A novel Regression-based Nearest Neighbour (RBNN) Classifier is developed for robust plug load identification under varying voltages. Regression technique is used to recognize the behavior of plug loads based on their energy consumption signature and then use identified behavior for classification. The proposed algorithm is evaluated against the traditional classification algorithm on a dataset of 70 plug loads operating in varying voltage conditions. Experimental results show that the proposed algorithm performs better than standard classifiers in most of the cases. Also, a prototype smart power strip with the proposed load identification technique is designed. The remaining part of this thesis describes ways to improve event-based NILM techniques. An event-based NILM approach identifies ON or OFF events and extracts features to build appliance identification models. A study of event-based NILM algorithms under the influence of varying voltages is performed. Feature extraction and feature selection techniques are developed to reduce the impact of voltage variation. Appl

Abstract

In a blockchain network, full nodes aid in decentralization by validating incoming transactions and also by assisting new nodes who want to join the network. However, due to the high storage cost incurred by full nodes, the total number of full nodes in a system is falling. In order to curb this problem and maintain decentralization, we present Secure Raptor Encoder and Decoder (S-RED) built on Raptor Codes, which in addition to reducing storage costs, also has constant decoding time with respect to k (number of input blocks). We also present error-resilient double-peeling decoder that can recover original blocks with a high probability, even under the presence of adversaries. Blockchain popularity and application is not a new thing but blockchain is bounded by its physical capacity and is unable to accommodate the incoming traffic leading to high transaction confirmation latency. Hence a new data structure based on the direct acyclic graph is proposed popularly called blockdag which doesn’t limit blockchain physical capacity. Prism is a ledger based on blockdag technology, however, in Prism full nodes have to pay a high cost in terms of storage space due to the requirement of more blocks as compared to conventional blockchain. In order to reduce such high storage costs, we present Secure Luby Transform Raptor Encoder and Decoder (S-LTRED) built on LT and Raptor Codes. We present tri-peel decoder that can recover original blocks with a high probability, even under the presence of adversaries. S-LTRED shows to achieve a much lower bootstrap cost as compared to Prism and achieves this by reducing the storage consumption.