Rakshit Agrawal

Rakshit Agrawal
Camio

Doctor of Philosophy

About

25
Publications
3,542
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
122
Citations
Introduction
Rakshit Agrawal currently works at the Department of Computer Science , University of California, Santa Cruz. Rakshit does research in Machine Learning, Crowdsourcing and Graphs. Their most recent publication is 'A New Family of Neural Networks Provably Resistant to Adversarial Attacks'.

Publications

Publications (25)
Preprint
Malicious scripts are an important computer infection threat vector in the wild. For web-scale processing, static analysis offers substantial computing efficiencies. We propose the ScriptNet system for neural malicious JavaScript detection which is based on static analysis. We use the Convoluted Partitioning of Long Sequences (CPoLS) model, which p...
Preprint
Graph edges, along with their labels, can represent information of fundamental importance, such as links between web pages, friendship between users, the rating given by users to other users or items, and much more. We introduce LEAP, a trainable, general framework for predicting the presence and properties of edges on the basis of the local struct...
Preprint
Full-text available
Social networks offer a ready channel for fake and misleading news to spread and exert influence. This paper examines the performance of different reputation algorithms when applied to a large and statistically significant portion of the news that are spread via Twitter. Our main result is that simple crowdsourcing-based algorithms are able to iden...
Preprint
Adversarial attacks add perturbations to the input features with the intent of changing the classification produced by a machine learning system. Small perturbations can yield adversarial examples which are misclassified despite being virtually indistinguishable from the unperturbed input. Classifiers trained with standard neural network techniques...
Preprint
Malicious software, or malware, presents a continuously evolving challenge in computer security. These embedded snippets of code in the form of malicious files or hidden within legitimate files cause a major risk to systems with their ability to run malicious command sequences. Malware authors even use polymorphism to reorder these commands and cre...
Conference Paper
Full-text available
Fidgeting involves interacting with objects using repetitive hand movements. Before you can study its effects, you must first study the objects with which people choose to fidget. We present the findings of our five-phase three-month study with 28 children, 24 parents, and 2 teachers examining fidget material qualities and inherent interactions chi...
Preprint
Full-text available
Malicious scripts are an important computer infection threat vector. Our analysis reveals that the two most prevalent types of malicious scripts include JavaScript and VBScript. The percentage of detected JavaScript attacks are on the rise. To address these threats, we investigate two deep recurrent models, LaMP (LSTM and Max Pooling) and CPoLS (Co...
Article
Interactive systems have taken over the web and mobile space with increasing participation from users. Applications across every marketing domain can now be accessed through mobile or web where users can directly perform certain actions and reach a desired outcome. Actions of user on a system, though, can be representative of a certain intent. Abil...
Article
Full-text available
Many prediction problems can be phrased as inferences over local neighborhoods of graphs. The graph represents the interaction between entities, and the neighborhood of each entity contains information that allows the inferences or predictions. We present an approach for applying machine learning directly to such graph neighborhoods, yielding predi...
Conference Paper
In many collaborative systems it is useful to automatically estimate the quality of new contributions; the estimates can be used for instance to flag contributions for review. To predict the quality of a contribution by a user, it is useful to take into account both the characteristics of the revision itself, and the past history of contributions b...
Conference Paper
Full-text available
Following a disaster, there are generally many volunteers willing and able to help those affected. While volunteers can quickly contribute in terms of donations and aid collection, their ability to help further is limited by the complexity of tasks. While technologies exist for volunteers to be able to set up and assist in disaster response data co...
Conference Paper
Crowdsourcing and collaborative systems have become an important part of Computer Science system deployments. The research discussed in this paper designs and explores the use of collaborative systems for crowdsourced user participation in different kinds of tasks. Application focus of projects discussed here is mostly towards social good. This pap...
Conference Paper
In this publication, the authors describe KrishiEkta, an integrated knowledge distribution system for Indian Agriculture. KrishiEkta addresses the inadequate access of important information to various stakeholders in Indian agriculture. Access of rich internet platforms is not available to various farmers in small rural parts of the country. Krishi...
Article
In this paper the authors design, develop and test two Smartphone applications for Indian agriculture workers. The problem under concern necessitates the need to take into consideration the low literacy level of Indian farmers. The authors also present results from two surveys, one with the farmers and other with phone dealers catering to a majorit...
Conference Paper
In this paper, the authors design a sensor network architecture for soil and weather data collection and database generation. A multi-sensor device containing sensors for extracting soil characteristics operates as the first component of this architecture. It connects to a mobile phone which further connects to a base server system. An XML file con...
Article
In this publication, the authors present the concept to design a hand-held device for crop management (CMD). The proposed design is of a device capable of providing guidance to farmers with respect to irrigation, fertilization, scheduling of crop management and pest control. CMD incorporates modifications/additions in a low cost mobile phone conver...

Network

Cited By