Abhimanyu Roy

Bio: Abhimanyu Roy is an academic researcher from Institute of Management Technology, Ghaziabad. The author has contributed to research in topics: Deep learning & Payment. The author has an hindex of 3, co-authored 4 publications receiving 111 citations. Previous affiliations of Abhimanyu Roy include University of Virginia.

Deep learning detecting fraud in credit card transactions

Abstract: Credit card fraud resulted in the loss of $3 billion to North American financial institutions in 2017. The rise of digital payments systems such as Apple Pay, Android Pay, and Venmo has meant that loss due to fraudulent activity is expected to increase. Deep Learning presents a promising solution to the problem of credit card fraud detection by enabling institutions to make optimal use of their historic customer data as well as real-time transaction details that are recorded at the time of the transaction. In 2017, a study found that a Deep Learning approach provided comparable results to prevailing fraud detection methods such as Gradient Boosted Trees and Logistic Regression. However, Deep Learning encompasses a number of topologies. Additionally, the various parameters used to construct the model (e.g. the number of neurons in the hidden layer of a neural network) also influence its results. In this paper, we evaluate a subsection of Deep Learning topologies — from the general artificial neural network to topologies with built-in time and memory components such as Long Short-term memory — and different parameters with regard to their efficacy in fraud detection on a dataset of nearly 80 million credit card transactions that have been pre-labeled as fraudulent and legitimate. We utilize a high performance, distributed cloud computing environment to navigate past common fraud detection problems such as class imbalance and scalability. Our analysis provides a comprehensive guide to sensitivity analysis of model parameters with regard to performance in fraud detection. We also present a framework for parameter tuning of Deep Learning topologies for credit card fraud detection to enable financial institutions to reduce losses by preventing fraudulent activity.

Management of mobile health projects in developing countries: An empirical study

Abstract: This study examines the successful implementation of mobile health (mHealth) projects. Public health has become a cause for concern for many nations. Countries such as India among others have launched several initiatives in recent years with the aim to improve public health. However, these initiatives always face the problem of allocating scarce resources to a large number of patients. One method to overcome this issue is through the use of mHealth technologies that can provide a variety of health-related services anywhere, anytime in a cost-effective manner. However, mHealth too faces the same issues as other information and communication technology projects such as lack of infrastructure and usability issues. For this reason, despite its potential, most mHealth projects still remain in the pilot stage. The study analyses 54 mHealth projects conducted in developing nations between 2011 and 2016 across multiple databases. To maintain accuracy and objectivity of results a text-mining approach is followed. The results of the analysis are then used to create questions for structured interviews which are conducted with different stakeholders of an mHealth project operated in Ahmedabad, India. The daily operations of the system are also observed. The interviews and observations of the functioning of the system are then analyzed using Atlas.ti. The results of this analysis are reconciled with the results of the text-mining to create a framework that governs the implementation and operation of mHealth projects in developing nations. The framework found seven key areas that influenced a mHealth project — infrastructure, design, personnel, end-user, value, public policy and finances. Thereafter, a review of management and technology literature across these dimensions was conducted to develop recommendations for successful implementation of projects along these seven dimensions identified in the framework.

A feasibility study for the development of an employment system for underserved communities

Abstract: This study deals with the formulation of a technology-based solution to the problem of unemployment in underserved communities. Jobs for low-skilled and unskilled labor in the informal economy are acquired by means of referrals and contacts, but this approach suffers from making employment related information privileged only to a few individuals. In order to gauge the feasibility of a technology solution, the employment seeking process was studied through interviews conducted with all stakeholders - employers, candidates and intermediaries like NGOs and employment agencies. The studies were conducted at slum communities in Ahmedabad, India. After the interviews, two questionnaires were developed that drew from the interviews, one for the members of the community and the other for employers who recruited from the community. The employers also went through a semi-structured interview in order to capture any pertinent data that was not represented in the questionnaire. Based on the responses to the surveys a preliminary version of the system was constructed.

Poor Economics: A Radical Rethinking of the Way to Fight Global Poverty

Abstract: Poor Economics: A Radical Rethinking of the Way to Fight Global Poverty, by Abhijit Banerjee and Esther Duflo. Jackson, TN: PublicAffairs, 2011. 320 pp. ISBN: 978-1-58648-798-0 (hbk.). US$26.99. Po...

The application of internet of things in healthcare: a systematic literature review and classification

Abstract: The Internet of Things (IoT) is an ecosystem that integrates physical objects, software and hardware to interact with each other. Aging of population, shortage of healthcare resources, and rising medical costs make IoT-based technologies necessary to be tailored to address these challenges in healthcare. This systematic literature review has been conducted to determine the main application area of IoT in healthcare, components of IoT architecture in healthcare, most important technologies in IoT, characteristics of cloud-based architecture, security and interoperability issues in IoT architecture and effects, and challenges of IoT in healthcare. Sixty relevant papers, published between 2000 and 2016, were reviewed and analyzed. This analysis revealed that home healthcare service was one of the main application areas of IoT in healthcare. Cloud-based architecture, by providing great flexibility and scalability, has been deployed in most of the reviewed studies. Communication technologies including wireless fidelity (Wi-Fi), Bluetooth, radio-frequency identification (RFID), ZigBee, and Low-Power Wireless Personal Area Networks (LoWPAN) were frequently used in different IoT models. The studies regarding the security and interoperability issues in IoT architecture in health are still low in number. With respect to the most important effects of IoT in healthcare, these included ability of information exchange, decreasing stay of hospitalization and healthcare costs. The main challenges of IoT in healthcare were security and privacy issues.

TinyML-Enabled Frugal Smart Objects: Challenges and Opportunities

Abstract: The TinyML paradigm proposes to integrate Machine Learning (ML)-based mechanisms within small objects powered by Microcontroller Units (MCUs) This paves the way for the development of novel applications and services that do not need the omnipresent processing support from the cloud, which is power consuming and involves data security and privacy risks In this work, a comprehensive review of the novel TinyML ecosystem is provided The related challenges and opportunities are identified and the potential services that will be enabled by the development of truly smart frugal objects are discussed As a main contribution of this paper, a detailed survey of the available TinyML frameworks for integrating ML algorithms within MCUs is provided Besides, aiming at illustrating the given discussion, a real case study is presented Concretely, we propose a multi-Radio Access Network (RAT) architecture for smart frugal objects The issue of selecting the most adequate communication interface for sending sporadic messages considering both the status of the device and the characteristics of the data to be sent is addressed To this end, several TinyML frameworks are evaluated and the performances of a number of ML algorithms embedded in an Arduino Uno board are analyzed The attained results reveal the validity of the TinyML approach, which successfully enables the integration of techniques such as Neural Networks (NNs), Support Vector Machine (SVM), decision trees, or Random Forest (RF) in frugal objects with constrained hardware resources The outcomes also show promising results in terms of algorithm's accuracy and computation performance

GEE: A Gradient-based Explainable Variational Autoencoder for Network Anomaly Detection

Abstract: This paper looks into the problem of detecting network anomalies by analyzing NetFlow records. While many previous works have used statistical models and machine learning techniques in a supervised way, such solutions have the limitations that they require large amount of labeled data for training and are unlikely to detect zero-day attacks. Existing anomaly detection solutions also do not provide an easy way to explain or identify attacks in the anomalous traffic. To address these limitations, we develop and present GEE, a framework for detecting and explaining anomalies in network traffic. GEE comprises of two components: (i)Variational Autoencoder (VAE)- an unsupervised deep-learning technique for detecting anomalies, and (ii)a gradient-based fingerprinting technique for explaining anomalies. Evaluation of GEE on the recent UGR dataset demonstrates that our approach is effective in detecting different anomalies as well as identifying fingerprints that are good representations of these various attacks.