-
Study of Wolf-Rayet stars using uGMRT
Authors:
Anindya Saha,
Anandmayee Tej,
Santiago del Palacio,
Michaël De Becker,
Paula Benaglia,
Ishwara Chandra CH,
Prachi Prajapati
Abstract:
In recent years, systems involving massive stars with large wind kinetic power have been considered as promising sites for investigating relativistic particle acceleration in low radio frequencies. With this aim, we observed two Wolf-Rayet systems, WR 114 and WR 142, using upgraded Giant Meterwave Radio Telescope observations in Band 4 (550-950 MHz) and Band 5 (1050-1450 MHz). None of the targets…
▽ More
In recent years, systems involving massive stars with large wind kinetic power have been considered as promising sites for investigating relativistic particle acceleration in low radio frequencies. With this aim, we observed two Wolf-Rayet systems, WR 114 and WR 142, using upgraded Giant Meterwave Radio Telescope observations in Band 4 (550-950 MHz) and Band 5 (1050-1450 MHz). None of the targets was detected at these frequencies. Based on the non-detection, we report 3$σ$ upper limits to the radio flux densities at 735 and 1260 MHz (123 and 66 $μ$Jy for WR 114, and 111 and 96 $μ$Jy for WR 142, respectively). The plausible scenarios to interpret this non-detection are presented.
△ Less
Submitted 24 June, 2024;
originally announced June 2024.
-
Optical polarisation study of Galactic Open clusters
Authors:
Namita Uppal,
Shashikiran Ganesh,
Santosh Joshi,
Mrinmoy Sarkar,
Prachi Prajapati,
Athul Dileep
Abstract:
Dust is a ubiquitous component in our Galaxy. It accounts for only $1\%$ mass of the ISM but still is an essential part of the Galaxy. It affects our view of the Galaxy by obscuring the starlight at shorter wavelengths and re-emitting in longer wavelengths. Studying the dust distribution in the Galaxy at longer wavelengths may cause discrepancies due to distance ambiguity caused by unknown Galacti…
▽ More
Dust is a ubiquitous component in our Galaxy. It accounts for only $1\%$ mass of the ISM but still is an essential part of the Galaxy. It affects our view of the Galaxy by obscuring the starlight at shorter wavelengths and re-emitting in longer wavelengths. Studying the dust distribution in the Galaxy at longer wavelengths may cause discrepancies due to distance ambiguity caused by unknown Galactic potential. However, another aspect of dust, i.e., the polarisation of the background starlight, when combined with distance information, will help to give direct observational evidence of the number of dust clouds encountered in the line of sight. We observed 15 open clusters distributed at increasing distances in three lines of sight using two Indian national facilities. The measured polarisation results used to scrutinize the dust distribution and orientation of the local plane of sky magnetic fields towards selected directions. The analysis of the stars observed towards the distant cluster King 8 cluster shows two foreground layers at a distance of $\sim 500$ pc and $\sim$ 3500 pc. Similar analysis towards different clusters also results in multiple dust layers.
△ Less
Submitted 16 November, 2023;
originally announced November 2023.
-
Search for particle acceleration in two massive Wolf-Rayet stars using uGMRT observations
Authors:
Anindya Saha,
Anandmayee Tej,
Santiago del Palacio,
Michaël De Becker,
Paula Benaglia,
Ishwara Chandra CH,
Prachi Prajapati
Abstract:
Large wind kinetic power of Wolf-Rayet (WR) stars make them ideal targets in low radio frequencies to search for non-thermal emission due to relativistic particle acceleration. In this paper, we present observations of two WR stars, WR 114 and WR 142, in Band 4 (550-950 MHz) and Band 5 (1050-1450 MHz) using the upgraded Giant Meterwave Radio Telescope (uGMRT). Neither star is detected in the obser…
▽ More
Large wind kinetic power of Wolf-Rayet (WR) stars make them ideal targets in low radio frequencies to search for non-thermal emission due to relativistic particle acceleration. In this paper, we present observations of two WR stars, WR 114 and WR 142, in Band 4 (550-950 MHz) and Band 5 (1050-1450 MHz) using the upgraded Giant Meterwave Radio Telescope (uGMRT). Neither star is detected in the observed frequency bands, nor extended emission associated with them. The upper limit to the free-free radio emission from the stellar wind enables us to constrain the mass-loss rate of WR 114 to $\lesssim \rm 10^{-5}\,M_{ \odot}\,yr^{-1}$; this is a factor three smaller than previously estimated using spectroscopic modelling. If we further assume that the WR stars are binaries, the non-detection of synchrotron emission from the putative wind collision region implies that the stars are either in very wide binary systems away from periastron, or that the stars are in close binary systems with an orbital separation $<70$ AU for WR 114 and $<20$ AU for WR 142. The non-detection of low-frequency radio emission from these two systems thus provides evidence that narrows their nature, though it does not rule them out as bonafide particle-accelerating colliding-wind binaries.
△ Less
Submitted 9 September, 2023;
originally announced September 2023.
-
MRI-GAN: A Generalized Approach to Detect DeepFakes using Perceptual Image Assessment
Authors:
Pratikkumar Prajapati,
Chris Pollett
Abstract:
DeepFakes are synthetic videos generated by swapping a face of an original image with the face of somebody else. In this paper, we describe our work to develop general, deep learning-based models to classify DeepFake content. We propose a novel framework for using Generative Adversarial Network (GAN)-based models, we call MRI-GAN, that utilizes perceptual differences in images to detect synthesize…
▽ More
DeepFakes are synthetic videos generated by swapping a face of an original image with the face of somebody else. In this paper, we describe our work to develop general, deep learning-based models to classify DeepFake content. We propose a novel framework for using Generative Adversarial Network (GAN)-based models, we call MRI-GAN, that utilizes perceptual differences in images to detect synthesized videos. We test our MRI-GAN approach and a plain-frames-based model using the DeepFake Detection Challenge Dataset. Our plain frames-based-model achieves 91% test accuracy and a model which uses our MRI-GAN framework with Structural Similarity Index Measurement (SSIM) for the perceptual differences achieves 74% test accuracy. The results of MRI-GAN are preliminary and may be improved further by modifying the choice of loss function, tuning hyper-parameters, or by using a more advanced perceptual similarity metric.
△ Less
Submitted 28 February, 2022;
originally announced March 2022.
-
A Comprehensive Study of the Young Cluster IRAS 05100+3723: Properties, Surrounding Interstellar Matter, and Associated Star Formation
Authors:
R. K. Yadav,
M. R. Samal,
E. Semenko,
A. Zavagno,
S. Vaddi,
P. Prajapati,
D. K. Ojha,
A. K. Pandey,
M. Ridsdill-Smith,
J. Jose,
S. Patra,
S. Dutta,
P. Irawati,
S. Sharma,
D. K. Sahu,
N. Panwar
Abstract:
We present a comprehensive multiwavelength investigation of a likely massive young cluster `IRAS 05100+3723' and its environment with the aim to understand its formation history and feedback effects. We find that IRAS 05100+3723 is a distant ($\sim$3.2 kpc), moderate mass ($\sim$500 \msun), young ($\sim$3 Myr) cluster with its most massive star being an O8.5V-type. From spectral modeling, we estim…
▽ More
We present a comprehensive multiwavelength investigation of a likely massive young cluster `IRAS 05100+3723' and its environment with the aim to understand its formation history and feedback effects. We find that IRAS 05100+3723 is a distant ($\sim$3.2 kpc), moderate mass ($\sim$500 \msun), young ($\sim$3 Myr) cluster with its most massive star being an O8.5V-type. From spectral modeling, we estimate the effective temperature and log $g$ of the star as $\sim$33,000 K and $\sim$3.8, respectively. Our radio continuum observations reveal that the star has ionized its environment forming an HII region of size $\sim$2.7 pc, temperature $\sim$5,700 K, and electron density $\sim$165 cm$^{-3}$. However, our large-scale dust maps reveal that it has heated the dust up to several parsecs ($\sim$10 pc) in the range 17$-$28 K and the morphology of warm dust emission resembles a bipolar HII region. From dust and $^{13}$CO gas analyses, we find evidences that the formation of the HII region has occurred at the very end of a long filamentary cloud around 3 Myr ago, likely due to edge collapse of the filament. We show that the HII region is currently compressing a clump of mass $\sim$2700 \msun at its western outskirts, at the junction of the HII region and filament. We observe several 70 $μ$m point sources of intermediate-mass and class 0 nature within the clump. We attribute these sources as the second generation stars of the complex. We propose that the star formation in the clump is either induced or being facilitated by the compression of the expanding HII region onto the inflowing filamentary material.
△ Less
Submitted 17 November, 2021;
originally announced November 2021.
-
An Empirical Analysis of Image-Based Learning Techniques for Malware Classification
Authors:
Pratikkumar Prajapati,
Mark Stamp
Abstract:
In this paper, we consider malware classification using deep learning techniques and image-based features. We employ a wide variety of deep learning techniques, including multilayer perceptrons (MLP), convolutional neural networks (CNN), long short-term memory (LSTM), and gated recurrent units (GRU). Amongst our CNN experiments, transfer learning plays a prominent role specifically, we test the VG…
▽ More
In this paper, we consider malware classification using deep learning techniques and image-based features. We employ a wide variety of deep learning techniques, including multilayer perceptrons (MLP), convolutional neural networks (CNN), long short-term memory (LSTM), and gated recurrent units (GRU). Amongst our CNN experiments, transfer learning plays a prominent role specifically, we test the VGG-19 and ResNet152 models. As compared to previous work, the results presented in this paper are based on a larger and more diverse malware dataset, we consider a wider array of features, and we experiment with a much greater variety of learning techniques. Consequently, our results are the most comprehensive and complete that have yet been published.
△ Less
Submitted 24 March, 2021;
originally announced March 2021.
-
VHE gamma-ray detection of FSRQ QSO B1420+326 and modeling of its enhanced broadband state in 2020
Authors:
V. A. Acciari,
S. Ansoldi,
L. A. Antonelli,
A. Arbet Engels,
M. Artero,
K. Asano,
D. Baack,
A. Babić,
A. Baquero,
U. Barres de Almeida,
J. A. Barrio,
J. Becerra González,
W. Bednarek,
L. Bellizzi,
E. Bernardini,
M. Bernardos,
A. Berti,
J. Besenrieder,
W. Bhattacharyya,
C. Bigongiari,
A. Biland,
O. Blanch,
G. Bonnoli,
Ž. Bošnjak,
G. Busetto
, et al. (209 additional authors not shown)
Abstract:
Context. QSO B1420+326 is a blazar classified as a Flat Spectrum Radio Quasar (FSRQ). In the beginning of 2020 it underwent an enhanced flux state. An extensive multiwavelength campaign allowed us to trace the evolution of the flare. Aims. We search for VHE gamma-ray emission from QSO B1420+326 during this flaring state. We aim to characterize and model the broadband emission of the source over di…
▽ More
Context. QSO B1420+326 is a blazar classified as a Flat Spectrum Radio Quasar (FSRQ). In the beginning of 2020 it underwent an enhanced flux state. An extensive multiwavelength campaign allowed us to trace the evolution of the flare. Aims. We search for VHE gamma-ray emission from QSO B1420+326 during this flaring state. We aim to characterize and model the broadband emission of the source over different phases of the flare. Methods. The source was observed with a number of instruments in radio, near infrared, optical (including polarimetry and spectroscopy), ultra-violet, X-ray and gamma-ray bands. We use dedicated optical spectroscopy results to estimate the accretion disk and the dust torus luminosity. We perform spectral energy distribution modeling in the framework of combined Synchrotron-Self-Compton and External Compton scenario in which the electron energy distribution is partially determined from acceleration and cooling processes. Results. During the enhanced state the flux of both SED components drastically increased and the peaks were shifted to higher energies. Follow up observations with the MAGIC telescopes led to the detection of very-high-energy gamma-ray emission from this source, making it one of only a handful of FSRQs known in this energy range. Modeling allows us to constrain the evolution of the magnetic field and electron energy distribution in the emission region. The gamma-ray flare was accompanied by a rotation of the optical polarization vector during a low polarization state. Also, a new, superluminal radio knot contemporaneously appeared in the radio image of the jet. The optical spectroscopy shows a prominent FeII bump with flux evolving together with the continuum emission and a MgII line with varying equivalent width.
△ Less
Submitted 21 December, 2020;
originally announced December 2020.
-
Mechanical aspects of Near-Infrared Imager Spectrometer and Polarimeter
Authors:
Prashanth Kumar Kasarla,
Pitamber Singh Patwal,
Hitesh Kumar L. Adalja,
Satya Narain Mathur,
Deekshya Roy Sarkar,
Alka Singh,
Archita Rai,
Prachi Vinod Prajapati,
Sachindra Naik,
Amish B. Shah,
Shashikiran Ganesh,
Kiran S. Baliyan
Abstract:
Near-infrared Imager Spectrometer and Polarimeter (NISP) is a camera, an intermediate resolution spectrograph and an imaging polarimeter being developed for upcoming 2.5m telescope of Physical Research Laboratory at Mount Abu, India. NISP is designed to work in the Near-IR (0.8-2.5 micron) using a H2RG detector. Collimator and camera lenses would transfer the image from the focal plane of the tele…
▽ More
Near-infrared Imager Spectrometer and Polarimeter (NISP) is a camera, an intermediate resolution spectrograph and an imaging polarimeter being developed for upcoming 2.5m telescope of Physical Research Laboratory at Mount Abu, India. NISP is designed to work in the Near-IR (0.8-2.5 micron) using a H2RG detector. Collimator and camera lenses would transfer the image from the focal plane of the telescope to the detector plane. The entire optics, mechanical support structures, detector-SIDECAR assembly will be cooled to cryo-temperatures using an open cycle Liquid Nitrogen tank inside a vacuum Dewar. GFRP support structures would be used to isolate cryogenic system from the Dewar. Two layer thermal shielding would be used to reduce the radiative heat transfer. Molecular sieve (getter) would be used to enhance the vacuum level inside Dewar. Magnet-reedswitch combination are used for absolute positioning of filterwheels. Here we describe the mechanical aspects in detail.
△ Less
Submitted 16 December, 2020;
originally announced December 2020.
-
Electronics design and development of Near-Infrared Imager, Spectrometer and Polarimeter
Authors:
Deekshya Roy Sarkar,
Amish B. Shah,
Alka Singh,
Pitamber Singh Patwal,
Prashanth Kumar Kasarla,
Archita Rai,
Prachi Vinod Prajapati,
Hitesh Kumar L. Adalja,
Satya N. Mathur,
Sachindra Naik,
Shashikiran Ganesh,
Kiran S. Baliyan
Abstract:
NISP, a multifaceted near-infrared instrument for the upcoming 2.5m IR telescope at MIRO Gurushikhar, Mount Abu, Rajasthan, India is being developed at PRL, Ahmedabad. NISP will have wide (FOV = 10' x 10') field imaging, moderate (R=3000) spectroscopy and imaging polarimetry operating modes. It is designed based on 0.8 to 2.5 micron sensitive, 2048 X 2048 HgCdTe (MCT) array detector from Teledyne.…
▽ More
NISP, a multifaceted near-infrared instrument for the upcoming 2.5m IR telescope at MIRO Gurushikhar, Mount Abu, Rajasthan, India is being developed at PRL, Ahmedabad. NISP will have wide (FOV = 10' x 10') field imaging, moderate (R=3000) spectroscopy and imaging polarimetry operating modes. It is designed based on 0.8 to 2.5 micron sensitive, 2048 X 2048 HgCdTe (MCT) array detector from Teledyne. Optical, Mechanical and Electronics subsystems are being designed and developed in-house at PRL. HAWAII-2RG (H2RG) detector will be mounted along with controlling SIDECAR ASIC inside LN2 filled cryogenic cooled Dewar. FPGA based controller for H2RG and ASIC will be mounted outside the Dewar at room temperature. Smart stepper motors will facilitate motion of filter wheels and optical components to realize different operating modes. Detector and ASIC temperatures are servo controlled using Lakeshore's Temperature Controller (TC) 336. Also, several cryogenic temperatures will be monitored by TC for health checking of the instrument. Detector, Motion and Temperature controllers onboard telescope will be interfaced to USB Hub and fiber-optic trans-receiver. Remote Host computer interface to remote end trans-receiver will be equipped with in-house developed GUI software to control all functionalities of NISP. Design and development aspects of NISP Electronics will be presented in this conference.
△ Less
Submitted 16 December, 2020;
originally announced December 2020.
-
EMPOL: an EMCCD based optical imaging polarimeter
Authors:
Shashikiran Ganesh,
Archita Rai,
Aravind K,
Alka Singh,
Prachi Vinod Prajapati,
Ashish Mishra,
Prashanth Kasarla,
Deekshya Roy Sarkar,
Pitamber Singh Patwal,
Namita Uppal,
Sunil Chandra,
Satyanarain Mathur,
Amish B Shah,
Kiran S Baliyan,
U C Joshi
Abstract:
An Andor 1K $\times$ 1K EMCCD detector has been used to develop an optical imaging polarimeter for use at the Cassegrain focus of 1.2 m telescope of PRL. The optics is derived from an older single-element detector instrument and consists of a rotating half-wave plate as modulator and a Foster prism as an analyser. The field of view of the instrument is 3 $\times$ 3 sq arcmin. We describe the instr…
▽ More
An Andor 1K $\times$ 1K EMCCD detector has been used to develop an optical imaging polarimeter for use at the Cassegrain focus of 1.2 m telescope of PRL. The optics is derived from an older single-element detector instrument and consists of a rotating half-wave plate as modulator and a Foster prism as an analyser. The field of view of the instrument is 3 $\times$ 3 sq arcmin. We describe the instrument and the observational methodology in this document. Extensive observations have been carried out with this instrument covering a large variety of sources e.g. near-Earth asteroids, comets, Lynds dark nebulae, open clusters and AGN such as blazars. In the current communication, we discuss some results from the initial calibration runs while the other results will be presented elsewhere.
△ Less
Submitted 16 December, 2020;
originally announced December 2020.
-
Optical aspects of Near-Infrared Imager Spectrometer and Polarimeter instrument (NISP)
Authors:
Archita Rai,
Shashikiran Ganesh,
Sukamal K. Paul,
Prashanth K. Kasarla,
Prachi V. Prajapati,
Deekshya R. Sarkar,
Alka Singh,
Pitambar S. Patwal,
Hitesh K L Adalja,
Satya N. Mathur,
Sachindra Naik,
Amish B. Shah,
Kiran S. Baliyan
Abstract:
As a Near-IR instrument to PRL's upcoming 2.5 m telescope, NISP is designed indigeniously at PRL to serve as a multifaceted instrument. Optical, Mechanical and Electronics subsystems are being designed and developed in-house at PRL. It will consist of imaging, spectroscopy and imaging-polarimetry mode in the wavelength bands Y, J, H, Ks i.e. 0.8 - 2.5 micron. The detector is an 2K x 2K H2RG (MCT)…
▽ More
As a Near-IR instrument to PRL's upcoming 2.5 m telescope, NISP is designed indigeniously at PRL to serve as a multifaceted instrument. Optical, Mechanical and Electronics subsystems are being designed and developed in-house at PRL. It will consist of imaging, spectroscopy and imaging-polarimetry mode in the wavelength bands Y, J, H, Ks i.e. 0.8 - 2.5 micron. The detector is an 2K x 2K H2RG (MCT) array detector from Teledyne, which will give a large FOV of 10' x 10' in the imaging mode. Spectroscopic modes with resolving power of R ~ 3000, will be achieved using grisms. Spectroscopy will be available in single order and a cross-dispersed mode shall be planned for simultaneous spectra. The instrument enables multi-wavelength imaging-polarimetry using Wedged-Double Wollaston (WeDoWo) prisms to get single shot Stokes parameters (I, Q, U) for linear polarisation simultaneously, thus increasing the efficiency of polarisation measurements and reducing observation time.
△ Less
Submitted 15 December, 2020;
originally announced December 2020.
-
On the Human Control of a Multiple Quadcopters with a Cable-suspended Payload System
Authors:
Pratik Prajapati,
Sagar Parekh,
Vineet Vashista
Abstract:
A quadcopter is an under-actuated system with only four control inputs for six degrees of freedom, and yet the human control of a quadcopter is simple enough to be learned with some practice. In this work, we consider the problem of human control of a multiple quadcopters system to transport a cable-suspended payload. The coupled dynamics of the system, due to the inherent physical constraints, is…
▽ More
A quadcopter is an under-actuated system with only four control inputs for six degrees of freedom, and yet the human control of a quadcopter is simple enough to be learned with some practice. In this work, we consider the problem of human control of a multiple quadcopters system to transport a cable-suspended payload. The coupled dynamics of the system, due to the inherent physical constraints, is used to develop a leader-follower architecture where the leader quadcopter is controlled directly by a human operator and the followers are controlled with the proposed Payload Attitude Controller and Cable Attitude Controller. Experiments, where a human operator flew a two quadcopters system to transport a cable-suspended payload, were conducted to study the performance of proposed controller. The results demonstrated successful implementation of human control in these systems. This work presents the possibility of enabling manual control for on-the-go maneuvering of the quadcopter-payload system which motivates aerial transportation in the unknown environments.
△ Less
Submitted 3 April, 2020;
originally announced April 2020.
-
Predictive analysis of Bitcoin price considering social sentiments
Authors:
Pratikkumar Prajapati
Abstract:
We report on the use of sentiment analysis on news and social media to analyze and predict the price of Bitcoin. Bitcoin is the leading cryptocurrency and has the highest market capitalization among digital currencies. Predicting Bitcoin values may help understand and predict potential market movement and future growth of the technology. Unlike (mostly) repeating phenomena like weather, cryptocurr…
▽ More
We report on the use of sentiment analysis on news and social media to analyze and predict the price of Bitcoin. Bitcoin is the leading cryptocurrency and has the highest market capitalization among digital currencies. Predicting Bitcoin values may help understand and predict potential market movement and future growth of the technology. Unlike (mostly) repeating phenomena like weather, cryptocurrency values do not follow a repeating pattern and mere past value of Bitcoin does not reveal any secret of future Bitcoin value. Humans follow general sentiments and technical analysis to invest in the market. Hence considering people's sentiment can give a good degree of prediction. We focus on using social sentiment as a feature to predict future Bitcoin value, and in particular, consider Google News and Reddit posts. We find that social sentiment gives a good estimate of how future Bitcoin values may move. We achieve the lowest test RMSE of 434.87 using an LSTM that takes as inputs the historical price of various cryptocurrencies, the sentiment of news articles and the sentiment of Reddit posts.
△ Less
Submitted 16 January, 2020;
originally announced January 2020.
-
Investigating Particle Acceleration in the Wolf-Rayet Bubble G2.4+1.4
Authors:
Prachi Prajapati,
Anandmayee Tej,
Santiago del Palacio,
Paula Benaglia,
Ishwara-Chandra CH,
Sarita Vig,
Samir Mandal,
Swarna Kanti Ghosh
Abstract:
The supersonic winds produced by massive stars carry a large amount of kinetic power. In numerous scenarios such winds have been proven to produce shocks in which relativistic particles are accelerated emitting non-thermal radiation. Here, we report the first detection of non-thermal emission from a single stellar bubble, G2.4+1.4, associated with a WO star. We observed this source with the uGMRT…
▽ More
The supersonic winds produced by massive stars carry a large amount of kinetic power. In numerous scenarios such winds have been proven to produce shocks in which relativistic particles are accelerated emitting non-thermal radiation. Here, we report the first detection of non-thermal emission from a single stellar bubble, G2.4+1.4, associated with a WO star. We observed this source with the uGMRT in Band 4 ($550-850$ MHz) and Band 5 ($1050-1450$ MHz). We present intensity and spectral index maps for this source that are consistent with synchrotron emission (average spectral index, $α= -0.83 \pm 0.10$). The fraction of the available kinetic wind power that is converted into cosmic ray acceleration is estimated to be of the order of a few per cent. This finding constitutes an observational breakthrough and gives new insight on the non-thermal physical processes taking place in the environments of isolated massive stars. In particular, our results show that non-runaway isolated massive stars are capable of accelerating relativistic particles and are therefore confirmed as sources of Galactic cosmic rays.
△ Less
Submitted 26 September, 2019;
originally announced September 2019.
-
Transfer Learning for Image-Based Malware Classification
Authors:
Niket Bhodia,
Pratikkumar Prajapati,
Fabio Di Troia,
Mark Stamp
Abstract:
In this paper, we consider the problem of malware detection and classification based on image analysis. We convert executable files to images and apply image recognition using deep learning (DL) models. To train these models, we employ transfer learning based on existing DL models that have been pre-trained on massive image datasets. We carry out various experiments with this technique and compare…
▽ More
In this paper, we consider the problem of malware detection and classification based on image analysis. We convert executable files to images and apply image recognition using deep learning (DL) models. To train these models, we employ transfer learning based on existing DL models that have been pre-trained on massive image datasets. We carry out various experiments with this technique and compare its performance to that of an extremely simple machine learning technique, namely, k-nearest neighbors (\kNN). For our k-NN experiments, we use features extracted directly from executables, rather than image analysis. While our image-based DL technique performs well in the experiments, surprisingly, it is outperformed by k-NN. We show that DL models are better able to generalize the data, in the sense that they outperform k-NN in simulated zero-day experiments.
△ Less
Submitted 20 January, 2019;
originally announced March 2019.
