Foreseer: Efficiently Forecasting Malware Event Series with Long Short-Term Memory

Gogineni, Kailash; Derasari, Preet; Venkataramani, Guru

doi:10.1109/seed55351.2022.00016

Cited by 4 publications

(1 citation statement)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Compared to working on the raw time series data, symbolizing time series with ABBA provides numerous benefits including but not limited to (1) dimensionality reduction, (2) smoothing and noise reduction, and (3) explainable feature discretization. The time series features extracted by ABBA enable fast time series forecasting (Elsworth & Güttel, 2020b), anomaly detection (Chen & Güttel, 2023;Elsworth & Güttel, 2020a), event prediction (Gogineni et al, 2022), classification (Nguyen & Ifrim, 2023;Taktak et al, 2024), and other data-driven tasks in time series analysis (Harris et al, 2021;Wang et al, 2023). An example illustration of an ABBA symbolization is shown in Figure 1.…”

mentioning

confidence: 99%

fABBA: A Python library for the fast symbolic approximation of time series

Chen,

Güttel

2024

JOSS

View full text Add to dashboard Cite

mentioning

confidence: 99%

fABBA: A Python library for the fast symbolic approximation of time series

Chen,

Güttel

2024

JOSS

View full text Add to dashboard Cite

Integrating AI-driven threat intelligence and forecasting in the cyber security exercise content generation lifecycle

Zacharis,

Katos,

Patsakis

2024

Int. J. Inf. Secur.

View full text Add to dashboard Cite

The escalating complexity and impact of cyber threats require organisations to rehearse responses to cyber-attacks by routinely conducting cyber security exercises. However, the effectiveness of these exercises is limited by the exercise planners’ ability to replicate real-world scenarios in a timely manner that is, most importantly, tailored to the training audience and sector impacted. To address this issue, we propose the integration of AI-driven sectorial threat intelligence and forecasting to identify emerging and relevant threats and anticipate their impact in different industries. By incorporating such automated analysis and forecasting into the design of cyber security exercises, organisations can simulate real-world scenarios more accurately and assess their ability to respond to emerging threats. Fundamentally, our approach enhances the effectiveness of cyber security exercises by tailoring the scenarios to reflect the threats that are more relevant and imminent to the sector of the targeted organisation, thereby enhancing its preparedness for cyber attacks. To assess the efficacy of our forecasting methodology, we conducted a survey with domain experts and report their feedback and evaluation of the proposed methodology.

show abstract