Physical Sciences and Mathematics Commons^™

Management And Security Of Iot Systems Using Microservices, Tharun Theja Kammara

Master's Projects

Devices that assist the user with some task or help them to make an informed decision are called smart devices. A network of such devices connected to internet are collectively called as Internet of Things (IoT). The applications of IoT are expanding exponentially and are becoming a part of our day to day lives. The rise of IoT led to new security and management issues. In this project, we propose a solution for some major problems faced by the IoT devices, including the problem of complexity due to heterogeneous platforms and the lack of IoT device monitoring for security and …

Proposing An Optimized Algorithm For Consolidating Electric-Powered Shared Scooters Into Hubs For Efficiently Managing Their Charging And Maintenance Operations, Ojen Goshtasb

Master's Projects

The use of vehicles other than ones containing combustion engines have been adopted significantly over the past few years and the direction it’s taking seems to be the future of urban transportation. The hottest vehicle of choice currently is the electric scooter. They are small and portable, fast, and less costly compared to getting in a cab from Lyft or Uber to get around town. The goal of this paper is to make a proposal to drive the creation of a safe, efficient system for these scooters’ management. This must be beneficial to all parties involved; the rider, non-riders, and …

Deep Visual Recommendation System, Raksha Sunil

Master's Projects

Recommendation system is a filtering system that predicts ratings or preferences that a user might have. Recommendation system is an evolved form of our trivial information retrieval systems. In this paper, we present a technique to solve new item cold start problem. New item cold start problem occurs when a new item is added to a shopping website like Amazon.com. There is no metadata for this item, no ratings and no reviews because it’s a new item in the system. Absence of data results in no recommendation or bad recommendations. Our approach to solve new item cold start problem requires …

Virtual Robot Climbing Using Reinforcement Learning, Ujjawal Garg

Master's Projects

Reinforcement Learning (RL) is a field of Artificial Intelligence that has gained a lot of attention in recent years. In this project, RL research was used to design and train an agent to climb and navigate through an environment with slopes. We compared and evaluated the performance of two state-of-the-art reinforcement learning algorithms for locomotion related tasks, Deep Deterministic Policy Gradients (DDPG) and Trust Region Policy Optimisation (TRPO). We observed that, on an average, training with TRPO was three times faster than DDPG, and also much more stable for the locomotion control tasks that we experimented. We conducted experiments and …

Pantry: A Macro Library For Python, Derek Pang

Master's Projects

Python lacks a simple way to create custom syntax and constructs that goes outside of its own syntax rules. A paradigm that allows for these possibilities to exist within languages is macros. Macros allow for a shorter set of syntax to expand into a longer set of instructions at compile-time. This gives the capability to evolve the language to fit personal needs.

Pantry, implements a hygienic text-substitution macro system for Python. Pantry achieves this through the introduction of an additional preparsing step that utilizes parsing and lexing of the source code. Pantry proposes a way to simply declare a pattern …

Intra-Exchange Cryptocurrency Arbitrage Bot, Eric Han

Master's Projects

Cryptocurrencies are defined as a digital currency in which encryption techniques are utilized to regulate generation of units of currency and verify the transfer of funds, independent of a central governing body such as a bank. Due to the large number of cryptocurrencies currently available, there inherently exists many price discrepancies due to market inefficiencies. Market inefficiencies occur when the price of assets do not reflect their true value. In fact, these types of pricing discrepancies exist in other financial markets, including fiat currency exchanges and stock exchanges. However, these discrepancies are more significant in the cryptocurrency domain due to …

Variations On A Theme: Using Amino Acid Sequences To Generate Music, Aaron Kosmatin

Master's Projects

In this project, we explore using a musical space to represent the properties of amino acids. We consider previous mappings and explore the limitations of these mappings. In this exploration, we will propose a new method of mapping into musical spaces that extends the properties that can be represented. For this work, we will use amino acid sequences as our example mapping. The amino acid properties we will use include mass, charge, structure, and hydrophobicity. Finally, we will show how the different musical properties can be compared for similarity.

Gradubique: An Academic Transcript Database Using Blockchain Architecture, Thinh Nguyen

Master's Projects

Blockchain has been widely adopted in the last few years even though it is in its infancy. The first well-known application built on blockchain technology was Bitcoin, which is a decentralized and distributed ledger to record crypto-currency transactions. All of the transactions in Bitcoin are anonymously transferred and validated by participants in the network. Bitcoin protocol and its operations are so reliable that technologists have been inspired to enhance blockchain technologies and deploy it outside of the crypto-currency world. The demand for private and non-crypto-currency solutions have surged among consortiums because of the security and fault tolerant features of blockchain. …

Dynamic Hierarchical Cache Management For Cloud Ran And Multi- Access Edge Computing In 5g Networks, Deepika Pathinga Rajendiran

Master's Projects

Cloud Radio Access Networks (CRAN) and Multi-Access Edge Computing (MEC) are two of the many emerging technologies that are proposed for 5G mobile networks. CRAN provides scalability, flexibility, and better resource utilization to support the dramatic increase of Internet of Things (IoT) and mobile devices. MEC aims to provide low latency, high bandwidth and real- time access to radio networks. Cloud architecture is built on top of traditional Radio Access Networks (RAN) to bring the idea of CRAN and in MEC, cloud computing services are brought near users to improve the user’s experiences. A cache is added in both CRAN …

Image Segmentation And Classification Of Marine Organisms, Krishna Teja Vojjila

Master's Projects

To automate the arduous task of identifying and classifying images through their domain expertise, pioneers in the field of machine learning and computer vision invented many algorithms and pre-processing techniques. The process of classification is flexible with many user and domain specific alterations. These techniques are now being used to classify marine organisms to study and monitor their populations. Despite advancements in the field of programming languages and machine learning, image segmentation and classification for unlabeled data still needs improvement. The purpose of this project is to explore the various pre-processing techniques and classification algorithms that help cluster and classify …

Distinguishing Earthquakes And Noise Using Random Forest Algorithm, Nishita Narvekar

Master's Projects

Earthquakes are a major cause of life and property destruction. It is known that earthquakes radiate energy in the form of surface and body seismic waves. P-wave and S-waves are types of body waves. Both waves can be detected and recorded at an earthquake station. These waves can be analyzed to detect earthquakes. Most of the earthquake prediction techniques today are a combination of geophysics and signal processing, which are relatively complex. Machine learning can be used to learn the behavior of seismic waves and help in early detection. Machine learning can also be employed to process massive amounts of …

Outfit Recommender System, Nikita Ramesh

Master's Projects

The online apparel retail market size in the United States is worth about seventy-two billion US dollars. Recommendation systems on retail websites generate a lot of this revenue. Thus, improving recommendation systems can increase their revenue. Traditional recommendations for clothes consisted of lexical methods. However, visual-based recommendations have gained popularity over the past few years. This involves processing a multitude of images using different image processing techniques. In order to handle such a vast quantity of images, deep neural networks have been used extensively. With the help of fast Graphics Processing Units, these networks provide results which are extremely accurate, …

Genetic Barcode Identification With Profile Hidden Markov Models, Vishrut Sharma

Master's Projects

DNA barcoding is a method that uses an organism’s DNA to identify its species. The gene cytochrome c oxidase I (COI) has been used effectively as a DNA barcode to identify organisms and elucidate relationships among species [1]. There also exists a database BOLD (Barcode Of Life Database) that contains COI sequences used for DNA barcoding for more than 1 million different species. Using BOLD to identify samples that have a match in the database is an uncomplicated process. However, this method fails to determine samples that are absent from the database. Given a sample that is not represented in …

A Neural Network Classifier For The Coi Barcode Gene, Saurabh Marathe

Master's Projects

Mitochondrial Cytochrome C Oxidase subunit I (CO I – to be read as “see – oh one”) is a 658 base pair region in the gene encoding that is proposed as standard barcode for animals. Meaning, the CO I is a special region found in animal DNA that is studied to identify the species of the animal. Currently, there is an implementation of an algorithm called ARBitrator which identifies and extracts these CO I sequences from enormous genes database called GenBank. The ARBitrator is good at extracting the CO I sequences that have better specificity and accuracy as compared to …

Sentiment Analysis Using An Ensemble Of Feature Selection Algorithms, Manankumar Bhagat

Master's Projects

To determine the opinion of any person experiencing any services or buying any product, the usage of Sentiment Analysis, a continuous research in the field of text mining, is a common practice. It is a process of using computation to identify and categorize opinions expressed in a piece of text. Individuals post their opinion via reviews, tweets, comments or discussions which is our unstructured information. Sentiment analysis gives a general conclusion of audits which benefit clients, individuals or organizations for decision making. The primary point of this paper is to perform an ensemble approach on feature reduction methods identified with …

Subtopics In Yelp Reviews, Riya Suchdev

Master's Projects

Yelp is a review platform that connects people to local businesses. It is a very popular platform that helps customers decide which business to choose. It relies on crowd sourced plain text reviews. From the business’s description some facts can be determined, such as category and location. However, more detailed description can be extracted from the reviews. Discovering latent topics and subtopics in Yelp reviews, can help summarize the reviews to gain knowledge. For example, we can deduce that reviews related to the Restaurant category tend to emphasize on service, food, order etc. Additionally, one can deduce positive or negative …

Feature Selection Using Genetic Algorithms, Vandana Kannan

Master's Projects

With the large amount of data of different types that are available today, the number of features that can be extracted from it is huge. The ever-increasing popularity of multimedia applications, has been a major factor for this, especially in the case of image data. Image data is used for several applications such as classification, retrieval, object recognition, and annotation. Often, utilizing the entire feature set for each of these activities can be not only be time consuming but can also negatively impact the performance. Given the large number of features, it is difficult to find the subset of features …

Validating Key-Value Based Implementations Of The Raft Consensus Algorithm For Distributed Systems, Deepthi Vishwanath

Master's Projects

Distributed systems are a group of systems connected via a network, all working towards achieving a common goal. To achieve fault tolerance and reliability, all the systems should work towards achieving consensus. Paxos is the most widely used consensus algorithm since 2 or 3 decades, but the shift is now happening towards a new algorithm known as Raft. Raft is a consensus algorithm (paper published in the year 2014) which is easier to understand and works like Paxos in terms of fault tolerance and performance. Since Raft is new, there is a need for a tool that verifies systems built …

Facial Emotion Recognition Using Machine Learning, Nitisha Raut

Master's Projects

Face detection has been around for ages. Taking a step forward, human emotion displayed by face and felt by brain, captured in either video, electric signal (EEG) or image form can be approximated. Human emotion detection is the need of the hour so that modern artificial intelligent systems can emulate and gauge reactions from face. This can be helpful to make informed decisions be it regarding identification of intent, promotion of offers or security related threats. Recognizing emotions from images or video is a trivial task for human eye, but proves to be very challenging for machines and requires many …

A Medical Price Prediction System, Anuja Tike

Master's Projects

The health care costs constitute a significant fraction of the U.S. economy. Nearly 20% of the Gross Domestic Product (GDP) is spent on health care. The health spending in the US is the highest among all developed nations in absolute numbers as well as a percentage of the economy. The U.S. government bears a large portion of seniors’ health expenditure through its Medicare program. The growing health related expenses combined with the fact that the baby-boomer generation is retiring, and hence they will be eligible for Medicare, puts a great burden on the U.S. exchequer. Therefore, it is essential to …

Vgm-Rnn: Recurrent Neural Networks For Video Game Music Generation, Nicolas Mauthes

Master's Projects

The recent explosion of interest in deep neural networks has affected and in some cases reinvigorated work in fields as diverse as natural language processing, image recognition, speech recognition and many more. For sequence learning tasks, recurrent neural networks and in particular LSTM-based networks have shown promising results. Recently there has been interest – for example in the research by Google’s Magenta team – in applying so-called “language modeling” recurrent neural networks to musical tasks, including for the automatic generation of original music. In this work we demonstrate our own LSTM-based music language modeling recurrent network. We show that it …

Android De-Shredder App, Vasudha Venkatesh

Master's Projects

Sensitive documents are usually shredded into strips before discarding them. Shredders are used to cut the pages of a document into thin strips of uniform thickness. Each shredded piece in the collection bin could belong to any of the pages in a document. The task of document reconstruction involves two steps: Identifying the page to which each shred belongs and rearranging the shreds within the page to their original position. The difficulty of the reconstruction process depends on the thickness of the shred and type of cut (horizontal or vertical). The thickness of the shred is directly proportional to the …

Using Filters In Time-Based Movie Recommender Systems, Ravee Khandagale

Master's Projects

On a very high level, a movie recommendation system is one which uses data about the user, data about the movie and the ratings given by a user in order to generate predictions for the movies that the user will like. This prediction is further presented to the user as a recommendation. For example, Netflix uses a recommendation system to predict movies and generate favorable recommendations for users based on their profiles and the profiles of users similar to them. In user-based collaborative filtering algorithm, the movies rated highly by the similar users of a particular user are considered as …

Joint Computation Offloading And Prioritized Scheduling In Mobile Edge Computing, Lingfang Gao

Master's Projects

With the rapid development of smart phones, enormous amounts of data are generated and usually require intensive and real-time computation. Nevertheless, quality of service (QoS) is hardly to be met due to the tension between resourcelimited (battery, CPU power) devices and computation-intensive applications. Mobileedge computing (MEC) emerging as a promising technique can be used to copy with stringent requirements from mobile applications. By offloading computationally intensive workloads to edge server and applying efficient task scheduling, energy cost of mobiles could be significantly reduced and therefore greatly improve QoS, e.g., latency. This paper proposes a joint computation offloading and prioritized task …

Machine Learning Playground, Adil Khan

Master's Projects

Machine learning is a science that “learns” about the data by finding unique patterns and relations in the data. There are a lot of libraries or tools available for processing machine learning datasets. You can upload your dataset in seconds and quickly start using these tools to get prediction results in a few minutes. However, generating an optimal model is a time consuming and tedious task. The tunable parameters (hyper-parameters) of any machine learning model may greatly affect the accuracy metrics. While most of the tools have models with default parameter setting to provide good results, they can often fail …

Anomaly Detection For Application Log Data, Aarish Grover

Master's Projects

In software development, there is an absolute requirement to ensure that a system once developed, functions at its best throughout its lifetime. Application log data is critical to maintaining application performance and thus techniques to parse, understand and detect anomalies in application log data are critical to ensuring efficiency in software development. While initially hampered by limited hardware and lack of quality datasets, anomaly detection techniques have recently received a surge of interest with advancements in machine learning technology and especially neural networks. In this paper, we explore anomaly detection, historical techniques to detect anomalies and recent advancements in neural …

A Convolutional Neural Network Based Approach For Visual Question Answering, Lavanya Abhinaya Koduri

Master's Projects

Computer Vision is a scientific discipline which involves the development of an algorithmic basis for the construction of intelligent systems that aim at analysis, understanding and extraction of useful information from visual data. This visual data can be plain images, video sequences, views from multiple cameras, etc. Natural Language Processing (NLP), is the ability of machines to read and understand human languages. Visual Question Answering (VQA), is a multi-discipline Artificial Intelligence (AI) research problem, which is a combination of Natural Language Processing (NLP), Computer Vision (CV), and Knowledge Reasoning (KR). Given an image and a question related to the image …

Analyzing Android Adware, Supraja Suresh

Master's Projects

Most Android smartphone apps are free; in order to generate revenue, the app developers embed ad libraries so that advertisements are displayed when the app is being used. Billions of dollars are lost annually due to ad fraud. In this research, we propose a machine learning based scheme to detect Android adware based on static and dynamic features. We collect static features from the manifest file, while dynamic features are obtained from network traffic. Using these features, we initially classify Android applications into broad categories (e.g., adware and benign) and then further classify each application into a more specific family. …

Sql Injection Detection Using Machine Learning Techniques And Multiple Data Sources, Kevin Ross

Master's Projects

SQL Injection continues to be one of the most damaging security exploits in terms of personal information exposure as well as monetary loss. Injection attacks are the number one vulnerability in the most recent OWASP Top 10 report, and the number of these attacks continues to increase. Traditional defense strategies often involve static, signature-based IDS (Intrusion Detection System) rules which are mostly effective only against previously observed attacks but not unknown, or zero-day, attacks. Much current research involves the use of machine learning techniques, which are able to detect unknown attacks, but depending on the algorithm can be costly in …

Modeling Human Migration Dynamics In Netlogo, Vikram Deshmukh

Master's Projects

Human Migration has often been the catalyst for the rise and fall of civilizations. It is imperative to study human migration dynamics if one is to gain insights into migratory behavior among human beings and how migration affects societies. There has been considerable research to study migration. This has given rise to some popular migration theories like the neoclassical approach, network migration, pull-push migration, etc. These theories shed light on some peculiar behaviors that influence the migration decision of an individual or a group, while also trying to predict the outcome of such actions. The goal of this project is …