When building a unified vision system or gradually adding new capabilities to a system, the usual assumption is that training data for all tasks is always available. However, as the number of tasks grows, storing and retraining on such data becomes infeasible. A new problem arises where we add new capabilities to a Convolutional Neural Network (CNN), but the training data for its existing capabilities are unavailable. We propose our Learning without Forgetting method, which uses only new task data to train the network while preserving the original capabilities. Our method performs favorably compared to commonly used feature extraction and fine-tuning adaption techniques and performs similarly to multitask learning that uses original task data we assume unavailable. A more surprising observation is that Learning without Forgetting may be able to replace fine-tuning with similar old and new task datasets for improved new task performance.

Learning without Forgetting

http://www.parkjonghyuk.net/lecture/2015-1st-lecture/networksecurity/Paper2.pdf

Security in cloud computing

In the last few years, the appealing features of cloud computing have been fueling the integration of cloud environments in the industry, which has been consequently motivating the research on related technologies by both the industry and the academia. The possibility of paying-as-you-go mixed with an on-demand elastic operation is changing the enterprise computing model, shifting on-premises infrastructures to off-premises data centers, accessed over the Internet and managed by cloud hosting providers. Regardless of its advantages, the transition to this computing paradigm raises security concerns, which are the subject of several studies. Besides of the issues derived from Web technologies and the Internet, clouds introduce new issues that should be cleared out first in order to further allow the number of cloud deployments to increase. This paper surveys the works on cloud security issues, making a comprehensive review of the literature on the subject. It addresses several key topics, namely vulnerabilities, threats, and attacks, proposing a taxonomy for their classification. It also contains a thorough review of the main concepts concerning the security state of cloud environments and discusses several open research topics.

/pdf/security-issues-in-cloud-environments-a-survey-1t9z0if3y4.pdf

Security issues in cloud environments: a survey

According to one embodiment, a computerized method comprises operations of instantiating a first virtual machine instance and a second virtual machine instance to run concurrently with the first virtual machine instance. The first virtual machine instance provides a first virtual operating environment while the second virtual machine instance is adapted to share the resources allocated to the first virtual machine instance. The second virtual machine instance is further adapted to allocate additional resources upon conducting a Copy-On Write operation.

Optimized resource allocation for virtual machines within a malware content detection system

DDoS attacks in cloud computing

This article proposes and analyzes a general cloud-based security overlay network that can be used as a transparent overlay network to provide services such as intrusion detection systems, antivirus and antispam software, and distributed denial-of-service prevention. The authors analyze each of these in-cloud security services in terms of resiliency, effectiveness, performance, flexibility, control, and cost.

Using Cloud Computing to Implement a Security Overlay Network

Enterprises are seeking toward the cloud horizon to expand their premises facilities. It provides several services in the market, such as Infrastructure as a Service (IaaS), Platform as a Service (PaaS) and Software as a Service (SaaS). This paper will discuss challenges regarding three information security concerns: confidentiality, integrity and availability. Most of the organizations are very much concerned about the ownership of their data. This paper will not only address security challenges for cloud computing including Identity and Access Management (IAM) but also present the current state authentication, authorization and auditing of users accessing the cloud along with emerging IAM protocols and standards.

Cloud computing security management

Cloud computing and digital forensics are emerging elds of technology. Unlike traditional digital forensics where the target environment can be almost completely isolated, acquired and can be under the investigators control; in cloud environments, the distribution of computation and storage poses unique and complex challenges to the investigators. Recently, the term \cloud forensics" has an increasing presence in the eld of digital forensics. In this state-of-the-art review, we included the most recent research eorts that used \cloud forensics" as a keyword and then classify the literature into three dimensions: (1) survey-based, (2) technology-based and (3) forensics-procedural-based. We discuss widely accepted standard bodies and their eorts to address the current trend of cloud forensics. Our aim is not only to reference related work based on the discussed dimensions, but also to analyse them and generate a mind map that will help in identifying research gaps. Finally, we summarize existing digital forensics tools and the available simulation environments that can be used for evidence acquisition, examination and cloud forensics test purposes.

/pdf/a-state-of-the-art-review-of-cloud-forensics-3h2lya56dw.pdf

A state-of-the-art review of cloud forensics

Cloud computing and digital forensics are both developing topics and researching these topics requires an understanding of the main aspects of both cloud computing and digital forensics. In cloud computing it is necessary not only to understand its characteristics and the different services and deployment models but also to survey the underpinning elements of cloud computing such as virtualization and the distributed computing which are important to identify its impact on current digital forensics guidelines and procedures. Unlike papers discussing the challenges and opportunities presented by cloud computing in relation to digital forensics, in this paper, we will discuss the underpinning cloud computing elements which are required to provide forensics friendly cloud services. Furthermore, we suggest a set of questions that will aid in the process of cloud forensics analysis.

Cloud forensics: A research perspective

In a heterogeneous and volatile environment such as the cloud, recording the underlying infrastructure state - snapshot-is an important paradigm for analysis, testing and verification of properties associated with distributed executions. In case of crime or e-discovery investigation, possible execution might require to generate digital forensics evidences. The aim of this research is to generate distributed snapshot and extract evidences in a forensically sound manner. The paper will cover literature review outcomes, preliminary findings, research methodology and work in progress.

Sameera Almulla

Papers

Using Cloud Computing to Implement a Security Overlay Network

Cloud computing security management

A state-of-the-art review of cloud forensics

Cloud forensics: A research perspective

A Distributed Snapshot Framework for Digital Forensics Evidence Extraction and Event Reconstruction from Cloud Environment