106 research outputs found
DDoS detection and mitigation using machine learning
No full textDistributed Denial of Service (DDoS) attacks are very common nowadays. It is evident that the current industry solutions, such as completely relying on the In- ternet Service Provider (ISP) or setting up a DDoS defense infrastructure, are not sufficient in detecting and mitigating DDoS attacks, hence consistent research is needed. In this thesis we first tried to understand how DDoS attacks happen, then we discussed a way to detect DDoS attacks using machine learning tools at the routers, instead of setting up a centralized analysis system. We have proposed a standard communication architecture which can be used across all the networking devices for mitigating DDoS attacks. We have also created a simulation program to demonstrate our detection technique.M.S.Includes bibliographical referencesby Arpit Ramesh Gawand
Empirical Study of Cloud Deployment Strategies: Guiding the choice between Containerization, Traditional and Hybrid Deployment
No full textThe empirical study provides a comprehensive evaluation of cloud deployment strategies – containerization, traditional virtual machines (VMs), and hybrid methods for three application types like static web applications, database web application and multithreaded applications with RabbitMQ. Motivated by the need for practical, data-driven guidance for cloud practitioners, the study evaluates key metrics such as performance, scalability, cost, reliability, and operational complexity. The findings shows that containerized deployment offer better performance and scalability for static web applications, hybrid deployments excel in performance, scalability and reliability for database web applications and multithreaded applications but both deployment strategies require complex setups which increases the operational complexity. While traditional VM deployments offer easy setup and low-cost offering usability for smaller applications and applications which do not have much load like academic projects or proof of concepts. A decision tree-based recommendation tool was developed to support practitioners in selecting appropriate deployment strategies based on the empirical data. Despite some of the limitations, including short evaluation period and resource constraints on scalability tests, this study shows a direction for future research in long performance analysis, broader application types, in depth scalability test and enhancing the recommendation tool. This future work will also help in commercializing this research study by the support of recommendation tool. The research ultimately provides actionable insights and practical tools to optimize cloud deployment strategies for its users, to ensure informed decision-making based on application requirements and scenarios
DISTRIBUTED BIOGEOGRAPHY BASED OPTIMIZATION FOR MOBILE ROBOTS
No full textI present hardware testing of an evolutionary algorithm (EA) known as distributed biogeography based optimization (DBBO). DBBO is an extended version of biogeography based optimization (BBO). Typically, EAs require a central computer to control the evaluation of candidate solutions to some optimization problem, and to control the sharing of information between those candidate solutions. DBBO, however, does not require a centralized unit to control individuals. Individuals independently run the EA and find a solution to a given optimization problem. Both BBO and DBBO are based on the theory of biogeography, which describes how organisms are distributed geographically in nature. I have compared the performance of BBO and DBBO by using fourteen benchmark functions that are commonly used to evaluate the performance of optimization algorithms. I perform both hardware and simulation experiments. Wall-following robots are used as hardware to implement the DBBO algorithm. Robots use two different controllers to maintain a constant distance from the wall: one is a proportional integral derivative (PID) controller and the other is a fuzzy controller. DBBO optimizes the performance of the robots with respect to the control parameters. During simulation experiments I used different EA mutation rates different staring points for the robots and different wheel bases. I have also done T-tests to analyze the statistical significance of performance differences and robustness tests to analyze the performance of the algorithms in the face of environmental changes. The results show that centralized BBO gives better optimization results than distributed BBO. DBBO gives less optimal solutions but it removes the necessity of centralized control. The results also show that the fuzzy controller performs better than the PID controlle
Elucidating the Effects of Interstitial Fluid Flow on Hepatocellular Carcinoma Invasion
No full textOver the last two decades with advancements in research, detection, and treatment of all cancer types in the United States, resulting in an overall 23% decrease in cancer related deaths, liver cancer has gone against this trend possessing an increased death rate. Globally, hepatocellular carcinoma (HCC), the most common form of liver cancer, ranks as the second leading cause of cancer related deaths with approximately 788,000 deaths annually. In recent years much emphasis has been placed on understanding the process of HCC cell invasion; however, it has become apparent that the progression of this disease is not solely dependent on just the cancer cells or biological factors, but also their interaction with the tumor microenvironment. A significant number of studies have shown that changes in biomechanical forces within the tumor microenvironment can alter cancer progression. Previous research has demonstrated that interstitial fluid flow (IFF), one of the biomechanical forces that is altered during tumor growth, can promote cancer cell invasion. The findings in this work elucidate the effects of IFF in HCC cell invasion. Using our 3D in vitro flow invasion assay, we demonstrate that IFF increases cellular invasion through autologous gradient formation of chemokines (CXCR4/CXCL12) that promote migration, a mechanism known as autologous chemotaxis. We also demonstrated that MEK/ERK signaling affects IFF-induced invasion; however, this pathway was separate from CXCR4/CXCL12 signaling. Increased matrix metalloproteinase (MMP) expression is a hallmark for cancer progression and poor prognosis. Biomechanical forces have been observed to increase the secretion of these proteolytic enzymes, which promote extracellular matrix degradation and tumor cell invasion. We observed an increase in MMP-9 and MMP-2 activity in HCC cells exposed to IFF. In total these findings indicate multiple mechanisms are at play in HCC flow-induced invasion, further emphasizing the significance biomechanical forces play in disease progression. Finally, by modifying our 3D in vitro flow invasion assay, we examined IFF in a relevant cell-based disease model where HCC cells are embedded in a stiff matrix. The increase in matrix stiffness is a result of tumor growth, shown to disturb the mechanical forces and biochemical signaling that occurs in the microenvironment, effectively promoting disease progression. HCC also possesses a very unique disease profile and risk factors; nearly 80% of HCCs occur from patients who suffer from chronic fibrosis or cirrhosis, where inflammation and hepatic wound-healing response attributes to the hepatocarcinogenesis. Many studies have observed cellular behavior of hepatocytes and HCC cells in a stiff matrix; however, none have observed the effect of IFF and a stiff microenvironment in HCC cells. The findings in this chapter confirm a synergistic relationship between IFF and matrix stiffness on HCC cell invasion. Ultimately the findings in this study provide a better foundational and mechanistic understanding of IFF and its effects on HCC cell invasion adding to the mounting evidence of how biomechanical forces in the tumor microenvironment influence cancer progression.Ph.D., Biomedical Engineering -- Drexel University, 201
Statistical field estimation and scale estimation for complex coastal regions and archipelagos
No full textThesis (S.M.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 2009.This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.Includes bibliographical references (p. 153-158).A fundamental requirement in realistic computational geophysical fluid dynamics is the optimal estimation of gridded fields and of spatial-temporal scales directly from the spatially irregular and multivariate data sets that are collected by varied instruments and sampling schemes. In this work, we derive and utilize new schemes for the mapping and dynamical inference of ocean fields in complex multiply-connected domains, study the computational properties of our new mapping schemes, and derive and investigate new schemes for adaptive estimation of spatial and temporal scales. Objective Analysis (OA) is the statistical estimation of fields using the Bayesian-based Gauss-Markov theorem, i.e. the update step of the Kalman Filter. The existing multi-scale OA approach of the Multidisciplinary Simulation, Estimation and Assimilation System consists of the successive utilization of Kalman update steps, one for each scale and for each correlation across scales. In the present work, the approach is extended to field mapping in complex, multiply-connected, coastal regions and archipelagos. A reasonably accurate correlation function often requires an estimate of the distance between data and model points, without going across complex land-forms. New methods for OA based on estimating the length of optimal shortest sea paths using the Level Set Method (LSM) and Fast Marching Method (FMM) are derived, implemented and utilized in general idealized and realistic ocean cases.(cont.) Our new methodologies could improve widely-used gridded databases such as the climatological gridded fields of the World Ocean Atlas (WOA) since these oceanic maps were computed without accounting for coastline constraints. A new FMM-based methodology for the estimation of absolute velocity under geostrophic balance in complicated domains is also outlined. Our new schemes are compared with other approaches, including the use of stochastically forced differential equations (SDE). We find that our FMM-based scheme for complex, multiply-connected, coastal regions is more efficient and accurate than the SDE approach. We also show that the field maps obtained using our FMM-based scheme do not require postprocessing (smoothing) of fields. The computational properties of the new mapping schemes are studied in detail. We find that higher-order schemes improve the accuracy of distance estimates. We also show that the covariance matrices we estimate are not necessarily positive definite because the Weiner Khinchin and Bochner relationships for positive deniteness are only valid for convex simply-connected domains. Several approaches to overcome this issue are discussed and qualitatively evaluated. The solutions we propose include introducing a small process noise or reducing the covariance matrix based on the dominant singular value decomposition.(cont.) We have also developed and utilized novel methodologies for the adaptive estimation of spatial-temporal scales from irregularly spaced ocean data. The three novel methodologies are based on the use of structure functions, short term Fourier transform and second generation wavelets. To our knowledge, this is the first time that adaptive methodologies for the spatial-temporal scale estimation are proposed. The ultimate goal of all these methods would be to create maps of spatial and temporal scales that evolve as new ocean data are fed to the scheme. This would potentially be a significant advance to the ocean community for better understanding and sampling of ocean processes.by Arpit Agarwal.S.M
Development of an astrocytic module for spiking neural networks on neuromorphic hardware
No full textAstrocytes have long been neglected in application to neuronal networks due to being electrically silent. While these glial cells have been hypothesized to serve as a support for neurons, recent research suggests that they may have a role in learning through spatial and temporal modulation of neurons. Astrocytes may form their own networks and communicate amongst themselves through calcium signaling. They have so far been absent in the Spiking neural networks (SNNs) and consequently, they have not been incorporated into neuromorphic chips such as Intel's Loihi. In this work, we discuss a new astrocytic module to extend the capabilities of Loihi to facilitate the inclusion of astrocytes in SNNs. This transformation from SNNs to Spiking Neural-Astrocytic Networks (SNANs) would enable researchers to both explore and leverage the capabilities of astrocytes in neuromorphic hardware. The module serves as a higher-level interface on top of Intel's NxSDK to allocate resources which serve as internal components of our astrocyte model to inject Slow Inward Current (SIC) and then introduce synchronous activity in the postsynaptic neurons. In addition, this work also addresses an additional project focused on the Unidimensional SLAM problem where we focus on solely the orientation of a robot placed in a variety of environments. We show that the spike-based algorithm implemented on Loihi requires approximately 100 times less power than the state-of-the-art GMapping algorithm implemented on a CPU. This work demonstrates the viability of Spiking Neural Networks running on Loihi as an alternative solution for mobile robots.M.S.Includes bibliographical reference
Elliptic Integral Approach to Large Deflection in Cantilever Beams: Theory and Validation
No full textIUPUIThis thesis investigates the large deflection behavior of cantilever beams under various
configurations and loading conditions. The primary objective is to uset an analytical model using
elliptic integrals to solve the second-order non-linear differential equations that govern the
deflection of these beams. The analytical model is implemented in Python and compared against
Finite Element Analysis (FEA) results obtained from ANSYS, ensuring the accuracy and
reliability of the model.
The study examines multiple beam configurations, including straight and inclined beams,
with both free and fixed tip slopes. Sensitivity analysis is conducted to assess the impact of key
parameters, such as Young’s modulus, beam height, width, and length, on the deflection behavior.
This analysis reveals critical insights into how variations in material properties and geometric
dimensions affect beam performance.
A detailed error analysis using Root Mean Square Error (RMSE) is performed to compare
the analytical model's predictions with the FEA results. The error analysis highlights any
discrepancies, demonstrating the robustness of the analytical approach. The results show that the
analytical model, based on elliptic integrals, closely matches the FEA results across a range of
configurations and loading scenarios.
The insights gained from this study can be applied to optimize the design of cantilever
beams in various engineering applications, including prosthetics, robotics, and structural
components.
Overall, this research provides a comprehensive understanding of the large deflection
behavior of cantilever beams and offers a reliable analytical tool for engineers to predict beam
performance under different conditions. The integration of Python-based numerical methods with
classical elliptic integral solutions presents a useful approach that enhances the precision and
applicability of beam deflection analysis
Improving a Reinforcement Learning Negotiating Agent’s Performance by Extracting Information from the Opponent’s Sequence of Offers
No full textWith the prospects of decentralized multi-agent systems becoming more prevalent in daily life, automated negotiation agents have made their place in these collaborative settings. They are an approach to promote communication between the agents in reaching solutions that are better for all involved.Recent literature has shown great potential in using machine learning, particularly model-free deep reinforcement learning like Proximal Policy Optimization (PPO), to develop more performant automated negotiation strategies. This work focuses on using information from the opponent's sequence of offers in a bilateral negotiation to further improve a baseline PPO agent. This involves extracting and representing information from the opponent's sequence of offers into a state vector with a fixed dimension to modify the input to the agent's policy, and then comparing the utilities this modified agent achieves to the baseline PPO agent. Since there is a large variety of numerical measures to represent a sequence of offers, an ablation study is conducted to investigate the effectiveness of each.The modified agents consistently reached solutions that had higher social welfare, although the agent's own utility did not improve or diminish significantly in comparison to the base PPO agent.https://github.com/brenting/negotiation_PPO The repository containing all the code this paper used. The code for this specific paper was done in the 'sequence-of-offers-single-thread' branch.CSE3000 Research ProjectComputer Science and Engineerin
Stock prediction, trading simulation and options volatility prediction using FASCOM++ (fuzzy associative cortical maps architecture)
No full textFuzzy Associative Cortical Maps Architecture (FASCOM) is inspired from the cortical maps found in many biological and artificial neural systems. The cortical maps organise and represent information obtained from sensory inputs and play important roles in learning and memory processes. FASCOM uses features inspired by the structure and functions of cortical maps and is integrated a linguistic fuzzy model to perform associative learning of input-output pairs. The project undertakes to improve the architecture of FASCOM to incorporate a learning mechanism, so that the network is capable of modifying its properties on the basis of the incoming data leading to better prediction and higher accuracy.
The author aims to validate the modified architecture of FASCOM by conducting benchmarking experiments and observing the improvement in the performance of the system over other systems. For this purpose, various classical datasets for classification and regression problems were used.
The author worked on many real-life application to observe FASCOM++’s performance on real-life data. One of the applications is stock data prediction where the author used Hong Kong stock data and predicted prices using FASCOM++ and compared the results with the actual prices. The analysis of FASCOM++’s performance helps in gauging its practical use in real-life applications such as stock trading. The author simulated a simple stock trading algorithm to compare and evaluate FASCOM++’s performance against other architectures.
The author explored other areas of applications and worked on options volatility prediction which is one of the core areas of research in the financial industry. By exploiting on the online learning capabilities FASCOM++ was able to perform better than the other architectures and demonstrated its capability to be a potential architecture for real-life purpose.Bachelor of Engineering (Computer Science
- …
