Large amounts of threat intelligence information about mal-ware attacks are available in disparate, typically unstructured, formats. Knowledge graphs can capture this information and its context using RDF triples represented by entities and relations. Sparse or inaccurate threat information, however, leads to challenges such as incomplete or erroneous triples. Named entity recognition (NER) and relation extraction (RE) models used to populate the knowledge graph cannot fully guaran-tee accurate information retrieval, further exacerbating this problem. This paper proposes an end-to-end approach to generate a Malware Knowledge Graph called MalKG, the first open-source automated knowledge graph for malware threat intelligence. MalKG dataset called MT40K1 contains approximately 40,000 triples generated from 27,354 unique entities and 34 relations. We demonstrate the application of MalKGin predicting missing malware threat intelligence information in the knowledge graph. For ground truth, we manually curate a knowledge graph called MT3K, with 3,027 triples generated from 5,741 unique entities and 22 relations. For entity prediction via a state-of-the-art entity prediction model(TuckER), our approach achieves 80.4 for the hits@10 metric (predicts the top 10 options for missing entities in the knowledge graph), and 0.75 for the MRR (mean reciprocal rank). We also propose a framework to automate the extraction of thousands of entities and relations into RDF triples, both manually and automatically, at the sentence level from1,100 malware threat intelligence reports and from the com-mon vulnerabilities and exposures (CVE) database.

"It will attempt to answer an age-old question that has eluded humanity for generations: whether life has ever existed elsewhere beyond our own planet," Thomas Zurbuchen, associate administrator for the space agency's science mission directorate in Washington, said of the Perseverance mission. Bristling with sensors, cameras, microphones and a robotic arm, the one-ton, SUV-size Perseverance rover is designed to look for rock or soil specimens that might harbor evidence of ancient life and pack what it finds into small tubes, to be cached for retrieval by future missions and brought back to Earth for analysis. NASA and the European Space Agency are discussing several mission scenarios that might return these samples by 2031, NASA officials said. If all goes well, NASA's Perseverance rover will join four previous robotic vehicles that pioneered exploration of the red planet. As the first rover on Mars, Sojourner was small and simple.

But some cities have stuck by the systems. In Detroit, where the police chief said the system was useful even though it almost never returned a perfect match without human guidance, city leaders last year approved further use of the software, saying it helped protect the public while empowering the police.

Humans have been trying and often failing to land robots on Mars since 1962. Four different space agencies from around the world have attempted this improbable feat 18 times. It's happened successfully only eight times--remarkably, all by NASA. On Thursday, the American space agency will try for number nine, as the Perseverance rover attempts to gently settle onto the Red Planet. Launched July 30, Perseverance carries an impressive array of instruments and technologies, including 23 cameras, three spectrometers, a radar mapper, and a weather station.

Computational tools for forecasting yields and prices for fresh produce have been based on traditional machine learning approaches or time series modeling. We propose here an alternate approach based on deep learning algorithms for forecasting strawberry yields and prices in Santa Barbara county, California. Building the proposed forecasting model comprises three stages: first, the station-based ensemble model (ATT-CNN-LSTM-SeriesNet_Ens) with its compound deep learning components, SeriesNet with Gated Recurrent Unit (GRU) and Convolutional Neural Network LSTM with Attention layer (Att-CNN-LSTM), are trained and tested using the station-based soil temperature and moisture data of Santa Barbara as input and the corresponding strawberry yields or prices as output. Secondly, the remote sensing ensemble model (SIM_CNN-LSTM_Ens), which is an ensemble model of Convolutional Neural Network LSTM (CNN-LSTM) models, is trained and tested using satellite images of the same county as input mapped to the same yields and prices as output. These two ensembles forecast strawberry yields and prices with minimal forecasting errors and highest model correlation for five weeks ahead forecasts. Finally, the forecasts of these two models are ensembled to have a final forecasted value for yields and prices by introducing a voting ensemble. Based on an aggregated performance measure (AGM), it is found that this voting ensemble not only enhances the forecasting performance by 5% compared to its best performing component model but also outperforms the Deep Learning (DL) ensemble model found in literature by 33% for forecasting yields and 21% for forecasting prices.

The application of remaining useful life (RUL) prediction has taken great importance in terms of energy optimization, cost-effectiveness, and risk mitigation. The existing RUL prediction algorithms mostly constitute deep learning frameworks. In this paper, we implement LSTM and GRU models and compare the obtained results with a proposed genetically trained neural network. The current models solely depend on Adam and SGD for optimization and learning. Although the models have worked well with these optimizers, even little uncertainties in prognostics prediction can result in huge losses. We hope to improve the consistency of the predictions by adding another layer of optimization using Genetic Algorithms. The hyper-parameters - learning rate and batch size are optimized beyond manual capacity. These models and the proposed architecture are tested on the NASA Turbofan Jet Engine dataset. The optimized architecture can predict the given hyper-parameters autonomously and provide superior results.

In open domain table-to-text generation, we notice that the unfaithful generation usually contains hallucinated content which can not be aligned to any input table record. We thus try to evaluate the generation faithfulness with two entity-centric metrics: table record coverage and the ratio of hallucinated entities in text, both of which are shown to have strong agreement with human judgements. Then based on these metrics, we quantitatively analyze the correlation between training data quality and generation fidelity which indicates the potential usage of entity information in faithful generation. Motivated by these findings, we propose two methods for faithful generation: 1) augmented training by incorporating the auxiliary entity information, including both an augmented plan-based model and an unsupervised model and 2) training instance selection based on faithfulness ranking. We show these approaches improve generation fidelity in both full dataset setting and few shot learning settings by both automatic and human evaluations.

Language EnglishTags:  Business ecosystems have expanded over the years owing to the many benefits of diverse, interconnected supply chains, prompting organizations to pursue close, collaborative relationships with their suppliers. However, this has led to increased cyber threats when organizations expose their networks to their supply chain and it only takes one supplier to have cybersecurity vulnerabilities to bring a business to its knees. To this point governments around the world have highlighted supply chains as an area for urgent attention in tackling cyber risk in the coming years. Looking Beyond Your Own Perimeter Over the last few years, many organizations have worked hard to improve their cyber defenses and are increasingly ‘harder targets’. However, for these well-defended organizations, now the greatest weaknesses in their defenses are their suppliers, who are typically less well defended but with whom they are highly interconnected. At the same time, the cyber threat landscape has intensified, and events of the past year have meant that security professionals are not only having to manage security in a remote working set up and ensure employees have good accessibility, they are also having to handle a multitude of issues from a distance while defending a much broader attack surface. As a result, points of vulnerability have become even more numerous, providing an attractive space for bad actors to disrupt and extort enterprises. Threats have escalated, including phishing and new variants of known threats, such as ransomware and Denial of Service (DDoS) attacks, as well as increases in supply chain attacks. But where supply chains are concerned, it is nearly impossible to effectively manage this risk unless you know the state of your suppliers’ defenses and continually ensure that they are comparable to your own. Organizations must deeply understand the cyber risks associated with the relationship and try to mitigate those risks to the degree possible. However, that’s easier said than done. With the sending and receiving of information essential for the supply chain to function, the only option is to better identify and manage the risks presented. This requires organizations to overhaul existing risk monitoring programs, technology investments and also to prioritize cyber and data security governance. Ensuring the Basics are in Place At the very least organizations should ensure that both they and their suppliers have the basic controls in place such as Cyber Essentials, NIST and ISO 27001, coupled with good data management controls. They should thoroughly vet and continuously monitor supply chain partners. They need to understand what data partners will need access to and why, and ultimately what level of risk this poses. Likewise, they need to understand what controls suppliers have in place to safeguard data and protect against incoming and outgoing cyber threats. This needs to be monitored, logged, and regularly reviewed and a baseline of normal activities between the organization and the supplier should be established. As well as effective processes, people play a key role in helping to minimize risk. Cybersecurity training should be given so that employees are aware of the dangers and know how to spot suspicious activity. They should be aware of data regulation requirements and understand what data can be shared with whom. And they should also know exactly what to do in the event of a breach, so a detailed incident response plan should be shared and regularly reviewed. IT best practices should be applied to minimize these risks. IT used effectively can automatically protect sensitive data so that when employees inevitably make mistakes, technology is there to safeguard the organization. Securely Transferring Information Between Suppliers So how do organizations transfer information between suppliers securely and how do they ensure that only authorized suppliers receive sensitive data? Here data classification tools are critical to ensure that sensitive data is appropriately treated, stored, and disposed of during its lifetime in accordance with its importance to the organization. Through appropriate classification, using visual labelling and metadata application to emails and documents, this protects the organization from the risk of sensitive data being exposed to unauthorized organizations further down the line through the supply chain. Likewise, data that isn’t properly encrypted in transit can be at risk of compromise, so using a secure and compliant mechanism for transferring data within the supply chain will significantly reduce risks. Managed File Transfer (MFT) software facilitates the automated sharing of data with suppliers. This secure channel provides a central platform for information exchanges and offers audit trails, user access controls, and other file transfer protections. Layering Security Defenses Organizations should also layer security defenses to neutralize any threats coming from a supplier. Due to its ubiquity, email is a particularly vulnerable channel and one that’s often exploited by cyber criminals posing as a trusted partner. Therefore, it is essential that organizations are adequately protected from incoming malware, embedded Advanced Persistent Threats, or any other threat that could pose a risk to the business. And finally, organizations need to ensure that documents uploaded and downloaded from the web are thoroughly analyzed, even if they are coming from a trusted source. To do this effectively, they need a solution that can remove risks from email, web and endpoints, yet still allows the transfer of information to occur. Adaptive Data Loss Prevention (DLP) allows the flow of information to continue while removing threats, protecting critical data, and ensuring compliance. It doesn’t become a barrier to business or impose a heavy management burden. This is important because traditional DLP ‘stop and block’ approaches have often resulted in too many delays to legitimate business communications and high management overheads associated with false positives. Cyber Criminal Attacks Set to Rise Many of the recent well publicized attacks have been nation state orchestrated. Going forward this is going to turn into criminal syndicate attacks. Cyber criminals already have the ransomware capabilities and now all they need to do is tie this up with targeting the supply chain. Therefore, making sure you have the right technologies, policies and training programs in place should be a top priority for organizations in 2021. If you are interested in finding out more about protecting your supply chain, download our Guide: “Managing Cybersecurity Risk in the Supply Chain.” Download the Guide Additional Resources On-Demand Webinar: Managing the Cybersecurity Supply Chain Risk in File TransfersTags: Featured: 1

In performative prediction, predictions guide decision-making and hence can influence the distribution of future data. To date, work on performative prediction has focused on finding performatively stable models, which are the fixed points of repeated retraining. However, stable solutions can be far from optimal when evaluated in terms of the performative risk, the loss experienced by the decision maker when deploying a model. In this paper, we shift attention beyond performative stability and focus on optimizing the performative risk directly. We identify a natural set of properties of the loss function and model-induced distribution shift under which the performative risk is convex, a property which does not follow from convexity of the loss alone. Furthermore, we develop algorithms that leverage our structural assumptions to optimize the performative risk with better sample efficiency than generic methods for derivative-free convex optimization.

The threat of adversarial examples has motivated work on training certifiably robust neural networks, to facilitate efficient verification of local robustness at inference time. We formalize a notion of global robustness, which captures the operational properties of on-line local robustness certification while yielding a natural learning objective for robust training. We show that widely-used architectures can be easily adapted to this objective by incorporating efficient global Lipschitz bounds into the network, yielding certifiably-robust models by construction that achieve state-of-the-art verifiable and clean accuracy. Notably, this approach requires significantly less time and memory than recent certifiable training methods, and leads to negligible costs when certifying points on-line; for example, our evaluation shows that it is possible to train a large tiny-imagenet model in a matter of hours. We posit that this is possible using inexpensive global bounds -- despite prior suggestions that tighter local bounds are needed for good performance -- because these models are trained to achieve tighter global bounds. Namely, we prove that the maximum achievable verifiable accuracy for a given dataset is not improved by using a local bound.