If you are in need of a lesson on just how much things have changed in the last 60 years, an anecdote from my childhood might suffice. My grandfather was a junk man, augmenting the income from his regular job by collecting scrap metal and selling it to metal recyclers. He knew the current scrap value of every common metal, and his garage and yard were stuffed with barrels of steel shavings, old brake drums and rotors, and miles of copper wire.

But his most valuable scrap was lead, specifically the weights used to balance car wheels, which he’d buy as waste from tire shops. The weights had spring steel clips that had to be removed before the scrap dealers would take them, which my grandfather did by melting them in a big cauldron over a propane burner in the garage. I clearly remember hanging out with him during his “melts,” fascinated by the flames and simmering pools of molten lead, completely unconcerned by the potential danger of the situation.

Fast forward a few too many decades and in an ironic twist I find myself living very close to the place where all that lead probably came from, a place that was also blissfully unconcerned by the toxic consequences of pulling this valuable industrial metal from tunnels burrowed deep into the Bitterroot Mountains. It didn’t help that the lead-bearing ores also happened to be especially rich in other metals including zinc and copper. But the real prize was silver, present in such abundance that the most productive silver mine in the world was once located in a place that is known as “Silver Valley” to this day. Together, these three metals made fortunes for North Idaho, with unfortunate side effects from the mining and refining processes used to win them from the mountains.

All Together Now

Thanks to the relative abundance of their ores and their physical and chemical properties, lead, silver, and zinc have been known and worked since prehistoric times. Lead, in fact, may have been the first metal our ancestors learned to smelt. It’s primarily the low melting points of these metals that made this possible; lead, for instance, melts at only 327°C, well within the range of a simple wood fire. It’s also soft and ductile, making it easy enough to work with simple tools that lead beads and wires dating back over 9,000 years have been found.

Unlike many industrial metals, minerals containing lead, silver, and zinc generally aren’t oxides of the metals. Rather, these three metals are far more likely to combine with sulfur, so their ores are mostly sulfide minerals. For lead, the primary ore is galena or lead (II) sulfide (PbS). Galena is a naturally occurring semiconductor, crystals of which lent their name to the early “crystal radios” which used a lump of galena probed with a fine cat’s whisker as a rectifier or detector for AM radio signals.

youtube.com/embed/-GxL13rid1w?…

Geologically, galena is found in veins within various metamorphic rocks, and in association with a wide variety of sulfide minerals. Exactly what minerals those are depends greatly on the conditions under which the rock formed. Galena crystallized out of low-temperature geological processes is likely to be found in limestone deposits alongside other sulfide minerals such as sphalerite, or zincblende, an ore of zinc. When galena forms under higher temperatures, such as those associated with geothermal processes, it’s more likely to be associated with iron sulfides like pyrite, or Fool’s Gold. Hydrothermal galenas are also more likely to have silver dissolved into the mineral, classifying them as argentiferous ores. In some cases, such as the mines of the Silver Valley, the silver is at high enough concentrations that the lead is considered the byproduct rather than the primary product, despite galena not being a primary ore of silver.

Like a Lead Bubble

How galena is extracted and refined depends on where the deposits are found. In some places, galena deposits are close enough to the surface that open-cast mining techniques can be used. In the Silver Valley, though, and in other locations in North America with commercially significant galena deposits, galena deposits follow deep fissures left by geothermal processes, making deep tunnel mining more likely to be used. The scale of some of the mines in the Silver Valley is hard to grasp. The galena deposits that led to the Bunker Hill stake in the 1880s were found at an elevation of 3,600′ (1,100 meters) above sea level; the shafts and workings of the Bunker Hill Mine are now 1,600′ (488 meters) below sea level, requiring miners to take an elevator ride one mile straight down to get to work.

Ore veins are followed into the rock using a series of tunnels or stopes that branch out from vertical shafts. Stopes are cut with the time-honored combination of drilling and blasting, freeing up hundreds of tons of ore with each blasting operation. Loose ore is gathered with a slusher, a bucket attached to a dragline that pulls ore back up the stope, or using mining loaders, low-slung payloaders specialized for operation in tight spaces.
Ore plus soap equals metal bubbles. Froth flotation of copper sulfide is similar to the process for extracting zinc sulfide. Source: Geomartin, CC BY-SA 4.0
Silver Valley galena typically assays at about 10% lead, making it a fairly rich ore. It’s still not rich enough, though, and needs to be concentrated before smelting. Most mines do the initial concentration on site, starting with the usual crushing, classifying, washing, and grinding steps. Ball mills are used to reduce the ore to a fine powder, mixed with water and surfactants to form a slurry, and pumped into a broad, shallow tank. Air pumped into the bottom of the tanks creates bubbles in the slurry that carry the fine lead particles up to the surface while letting the waste rock particles, or gangue, sink to the bottom. It seems counterintuitive to separate lead by floating it, but froth flotation is quite common in metal refining; we’ve seen it used to concentrate everything from lightweight graphite to ultradense uranium. It’s also important to note that this is not yet elemental lead, but rather still the lead sulfide that made up the bulk of the galena ore.

Once the froth is skimmed off and dried, it’s about 80% pure lead sulfide and ready for smelting. The Bunker Hill Mine used to have the largest lead smelter in the world, but that closed in 1982 after decades of operation that left an environmental and public health catastrophe in its wake. Now, concentrate is mainly sent to smelters located overseas for final processing, which begins with roasting the lead sulfide in a blast of hot air. This converts the lead sulfide to lead oxide and gaseous sulfur dioxide as a waste product:

After roasting, the lead oxide undergoes a reduction reaction to free up the elemental lead by adding everything to a blast furnace fueled with coke:

Any remaining impurities float to the top of the batch while the molten lead is tapped off from the bottom of the furnace.

Zinc!

A significant amount of zinc is also located in the ore veins of the Silver Valey, enough to become a major contributor to the district’s riches. The mineral sphalerite is the main zinc ore found in this region; like galena, it’s a sulfide mineral, but it’s a mixture of zinc sulfide and iron sulfide instead of the more-or-less pure lead oxide in galena. Sphalerite also tends to be relatively rich in industrially important contaminants like cadmium, gallium, germanium, and indium.
Most sphalerite ore isn’t this pretty. Source: Ivar Leidus, CC BY-SA 4.0.
Extraction of sphalerite occurs alongside galena extraction and uses mostly the same mining processes. Concentration also uses the froth flotation method used to isolate lead sulfide, albeit with different surfactants specific for zinc sulfide. Concentration yields a material with about 50% zinc by weight, with iron, sulfur, silicates, and trace metals making up the rest.

Purification of zinc from the concentrate is via a roasting process similar to that used for lead, and results in zinc oxide and more sulfur dioxide:

Originally, the Bunker Hill smelter just vented the sulfur dioxide out into the atmosphere, resulting in massive environmental damage in the Silver Valley. My neighbor relates his arrival in Idaho in 1970, crossing over the Lookout Pass from Montana on the then brand-new Interstate 90. Descending into the Silver Valley was like “a scene from Dante’s Inferno,” with thick smoke billowing from the smelter’s towering smokestacks trapped in the valley by a persistent inversion. The pine trees on the hillsides had all been stripped of needles by the sulfuric acid created when the sulfur dioxide mixed with moisture in the stale air. Eventually, the company realized that sulfur was too valuable to waste and started capturing it, and even built a fertilizer plant to put it to use. But the damage was done, and it took decades for the area to bounce back.

Recovering metallic zinc from zinc oxide is performed by reduction, again in a coke-fired blast furnace which collects the zinc vapors and condenses them to the liquid phase, which is tapped off into molds to create ingots. An alternative is electrowinning, where zinc oxide is converted to zinc sulfate using sulfuric acid, often made from the sulfur recovered from roasting. The zinc sulfate solution is then electrolyzed, and metallic zinc is recovered from the cathodes, melted, further purified if necessary, and cast into ingots.

Silver from Lead

If the original ore was argentiferous, as most of the Silver Valley’s galena is, now’s the time to recover the silver through the Parke’s process, a solvent extraction technique. In this case, the solvent is the molten lead, in which silver is quite soluble. The dissolved silver is precipitated by adding molten zinc, which has the useful property of reacting with silver while being immiscible with lead. Zinc also has a higher melting point than lead, meaning that as the temperature of the mixture drops, the zinc solidifies, carrying along any silver it combined with while in the molten state. The zinc-silver particles float to the top of the desilvered lead where they can be skimmed off. The zinc, which has a lower boiling point than silver, is driven off by vaporization, leaving behind relatively pure silver.

To further purify the recovered silver, cupellation is often employed. Cupellation is a pyrometallurgical process used since antiquity to purify noble metals by exploiting the different melting points and chemical properties of metals. In this case, silver contaminated with zinc is heated to the point where the zinc oxidizes in a shallow, porous vessel called a cupel. Cupels were traditionally made from bone ash or other materials rich in calcium carbonate, which gradually absorbs the zinc oxide, leaving behind a button of purified silver. Cupellation can also be used to purify silver directly from argentiferous galena ore, by differentially absorbing lead oxide from the molten solution, with the obvious disadvantage of wasting the lead:

Cupellation can also be used to recover small amounts of silver directly from refined lead, such as that in wheel weights:

youtube.com/embed/KoFAGEO5Fu8?…

If my grandfather had only known.

hackaday.com/2024/10/02/mining…

La scadenza si avvicina: il 18 ottobre entrerà in vigore la nuova Direttiva NIS2, una normativa che porterà cambiamenti radicali nel panorama della cybersicurezza in Europa. Le organizzazioni che non si adegueranno rischiano sanzioni salatissime, con multe che possono arrivare fino a 10 milioni di euro o il 2% del fatturato globale annuo, a seconda di quale importo sia maggiore. La domanda che tutte le aziende si pongono è: cos’è la NIS2 e come garantire di essere conformi entro la scadenza?

Cos’è la Direttiva NIS2 e perché è importante?

La Direttiva NIS2 (Network and Information Security) è stata introdotta dall’Unione Europea con l’obiettivo di rafforzare la resilienza e la sicurezza delle infrastrutture digitali critiche all’interno degli Stati Membri. Questa nuova regolamentazione, ufficialmente nota come Direttiva Europea 2022/2555, rappresenta un’evoluzione significativa rispetto alla sua predecessora, la NIS1, e affronta in modo più completo le sfide attuali poste dalla crescente minaccia di attacchi cibernetici su larga scala.

La direttiva stabilisce regole chiare e dettagliate per proteggere le infrastrutture vitali, richiedendo misure di sicurezza più rigorose e un livello di protezione più uniforme tra i paesi membri. Uno degli aspetti più innovativi della NIS2 è l’ampliamento del suo ambito di applicazione: ora vengono inclusi anche settori che prima non erano considerati cruciali, come l’industria chimica, la gestione dei rifiuti, e persino i servizi digitali e postali. Questo riflette l’importanza crescente che la sicurezza delle informazioni ha assunto in tutte le aree dell’economia digitale.

Le organizzazioni coinvolte sono suddivise in due gruppi principali:

• Soggetti importanti: questa categoria include una vasta gamma di settori come i servizi digitali, l’industria chimica, l’agroalimentare, la gestione dei rifiuti e i servizi postali, settori che pur non essendo fondamentali per la sopravvivenza immediata della società, sono comunque critici per il buon funzionamento delle economie e delle comunità.
• Soggetti essenziali: sono enti e aziende che operano in settori di importanza strategica come l’energia, la sanità, il settore finanziario, i trasporti, le infrastrutture digitali, le risorse idriche, lo spazio e la difesa.

Perché la conformità alla NIS2 non è negoziabile

La Direttiva NIS2 impone un sistema sanzionatorio molto severo per la mancata conformità. Le multe per i soggetti importanti possono raggiungere i 7 milioni di euro o l’1,4% del fatturato annuo globale. Per i soggetti essenziali, invece, le sanzioni arrivano fino a 10 milioni di euro o al 2% del fatturato. Inoltre, le organizzazioni che non ottemperano ai nuovi requisiti possono affrontare un significativo danno reputazionale, minando la fiducia dei clienti e degli investitori.

A differenza di altre normative, come il GDPR, la NIS2 impone controlli non solo sulla protezione dei dati personali, ma su tutto lo spettro delle infrastrutture IT, richiedendo alle organizzazioni di adottare misure preventive per garantire la continuità operativa e la resilienza contro incidenti e attacchi cibernetici.

Le principali misure per ottenere la conformità alla NIS2

Essere conformi alla NIS2 non significa semplicemente implementare qualche misura di sicurezza aggiuntiva. È necessario un approccio sistemico che coinvolga vari aspetti della gestione dei rischi IT. Tra le misure obbligatorie ci sono:

• Strategie di gestione del rischio cibernetico: identificare, valutare e mitigare i rischi legati alla sicurezza informatica.
• Protezione della supply chain: monitorare e proteggere l’intera catena di approvvigionamento da potenziali vulnerabilità.
• Continuità operativa e piani di ripristino: assicurarsi che le operazioni aziendali possano continuare anche in caso di un grave incidente.
• Gestione degli incidenti di sicurezza: creare piani di risposta rapida e strutturata in caso di attacchi o violazioni della sicurezza.
• Uso avanzato della crittografia e dell’autenticazione: adottare tecnologie come l’autenticazione a più fattori e la cifratura dei dati per proteggere informazioni sensibili.

Cubbit: una soluzione cloud innovativa per la compliance alla NIS2

Quando si tratta di garantire la conformità alla NIS2, l’adozione di soluzioni cloud sicure è fondamentale. Cubbit, una giovane e dinamica azienda italiana fondata a Bologna, sta rivoluzionando il mondo del cloud storage con il suo modello geo-distribuito.

A differenza dei fornitori di cloud tradizionali che concentrano i dati in pochi data center, Cubbit adotta un approccio completamente decentralizzato. I dati sono frammentati, crittografati e distribuiti su più nodi all’interno di un unico Paese, a scelta dell’utente, eliminando il rischio che la caduta di un singolo nodo comprometta la disponibilità dei dati.

Questa architettura garantisce una durabilità dei dati fino a 15 9, un livello di affidabilità diecimila volte superiore rispetto agli 11 9 offerti dalle soluzioni tradizionali. Inoltre, Cubbit offre un controllo senza precedenti sulla data residency, consentendo alle aziende di selezionare con precisione dove i loro dati vengono archiviati, in pieno rispetto delle normative NIS2 e GDPR.

Cubbit non si limita a offrire un’infrastruttura cloud sicura. Ha anche implementato avanzate tecnologie di protezione dai ransomware, come il versioning dei file e l’object lock, che impedisce la modifica o la cancellazione dei dati. Grazie a queste soluzioni, le organizzazioni possono recuperare versioni non compromesse dei file e proteggersi sia dagli attacchi malevoli che dagli errori umani.

Certificazioni di sicurezza e standard di eccellenza

Per offrire la massima tranquillità ai suoi clienti, Cubbit è soggetto a audit e certificazioni internazionali di terze parti, tra cui:

• ISO 9001:2015 per i sistemi di gestione della qualità
• ISO/IEC 27001:2013 per la sicurezza delle informazioni
• ISO/IEC 27017:2015 per la sicurezza nel cloud
• ISO/IEC 27018:2019 per la privacy e la protezione dei dati personali nel cloud
• Cybersecurity Made in Europe Label, una garanzia di qualità e affidabilità europea.

Cubbit è inoltre disponibile sulla piattaforma MePa e possiede la qualifica ACN (ex AgID), rendendolo idoneo per operare con enti pubblici e organizzazioni del settore privato che necessitano di livelli di sicurezza estremamente elevati.

Preparati alla NIS2 con Cubbit

Il tempo per adeguarsi alla NIS2 sta per scadere, e le conseguenze della non conformità possono essere devastanti. Cubbit rappresenta la soluzione ideale per le aziende che cercano un cloud sicuro, conforme e flessibile, capace di adattarsi ai nuovi requisiti normativi senza compromettere l’efficienza operativa.

Non aspettare oltre: visita il nostro sito web e scopri come Cubbit può supportare la tua organizzazione nella transizione verso la compliance NIS2, garantendo la protezione dei tuoi dati e la continuità delle operazioni aziendali.

L'articolo Guida alla conformità alla NIS2: La scadenza è a ottobre 2024 proviene da il blog della sicurezza informatica.

Artificial intelligence (AI) seems to be doing everything these days. Making images, making videos, and replacing most of us real human writers if you believe the hype. Maybe it’s all over! And yet, we persist, to write about yet another job taken over by AI: creating video games.

The research paper is entitled “Video Game Generation: A Practical Study using Mario.” The basic idea is whether a generative AI model can create an interactive video game by first training it on an existing game.

MarioVGG, as it is called, is a “text-to-video model.” It hasn’t built the Mario game that you’re familiar with, though. It takes player commands as text inputs—such as “run, or “jump”—and then outputs video frames showing the result in the ‘game.’ The model was trained on a dataset of frame-by-frame Super Mario Brothers game play, combined with data on user inputs at the time. The model shows an ability to generate believable video output for given player inputs, including basic game physics, item interactions, and collisions. It’s able to do this in a chained way, so that it can reasonably simulate a player making multiple actions and moving through a level of the game.

It’s not like playing a real Mario game yet, by any means. Regardless, the AI model has shown an ability to replicate the world of the game in a way that behaves relatively consistently with its established rules. If you’re in the field of video game development, though, you probably don’t have a lot to worry about just yet—you probably moved past making basic Mario clones years ago, so you’ve got quite an edge for now!

hackaday.com/2024/10/02/creati…

Introduction

In the ever-evolving landscape of cybersecurity, logs, that is information collected from various sources like network devices, endpoints, and applications, plays a crucial role in identifying and responding to threats. By analyzing this data, organizations can detect anomalies, pinpoint malicious activity, and mitigate potential cyberattacks before they cause significant damage. However, the sheer volume and complexity of logs often make them challenging to analyze effectively.

This is where machine learning (ML) comes into play. ML, a subset of artificial intelligence (AI), with its ability to process and analyze large datasets, offers a powerful solution to enhance threat detection capabilities. ML enables faster and more accurate identification of cyberthreats, helping organizations stay ahead of increasingly sophisticated attackers.

At Kaspersky, we have been using ML algorithms in our solutions for close on 20 years, and we have formulated ethical principles for the development and use of AI/ML. We utilize a variety of ML models and methods that are key to automating threat detection, anomaly recognition, and enhancing the accuracy of malware identification. In this post, we will share our experience hunting for new threats by processing Kaspersky Security Network (KSN) global threat data with ML tools to identify subtle new Indicators of Compromise (IoCs). We will also discuss challenges in implementing machine learning and interpreting threat hunting results.

The Kaspersky Security Network (KSN) infrastructure is designed to receive and process complex global cyberthreat data, transforming it into actionable threat intelligence that powers our products. A key source of threat-related data comes from voluntary contributions by our customers. To find out more, read about the principles of Kaspersky Security Network and our privacy policy.

The role of ML in global threat log analysis

Machine learning enables systems to learn from data and improve their performance over time without being explicitly programmed. When applied to cybersecurity logs, ML excels at analyzing large and complex datasets. It can automatically identify patterns, detect anomalies, and predict potential threats. When applied to our KSN global threat logs, ML can help predict and identify new threats.

One of the ML algorithms particularly well-suited for this task is Random Forest. This algorithm works by constructing multiple decision trees during training and assigning classes by the majority vote of the trees during scanning. Random Forest is especially effective in handling non-linear data, reducing the risk of overfitting, and providing insights into the importance of various features in the dataset. Random Forest is highly effective at identifying patterns, but this strength can lead to challenges in interpretability, particularly with larger models. It may produce positive results that are difficult to understand or explain due to the complexity of the decision-making process. Nevertheless, the benefits make it an excellent choice for analyzing logs and uncovering hidden cyberthreats.

ML Random Forest process

Reconstructing reality: the benefits of using ML in cybersecurity

Continuous learning allows ML models to detect subtle and novel cyberthreats, providing more robust defense. ML then acts to “reconstruct the cyber-reality” by transforming raw telemetry data into actionable insights that reflect the true state of a network or system’s security.

Reconstructing reality using logs: a depiction

The ability of ML-powered technology to analyze vast amounts of data in real time ensures that potential threats are identified and addressed more quickly, minimizing the window of vulnerability. Additionally, the adaptability of ML means that as new attack vectors emerge, models can still provide some level of protection without requiring updates, unlike traditional systems that often need constant rule revisions. This leads to a more proactive and effective cybersecurity posture, allowing organizations to stay ahead of increasingly sophisticated cyberattacks.

Ultimately, the ability of ML to partially reconstruct the cyber-reality from logs helps organizations stay ahead of cyberthreats by offering a clearer, more precise view of their security posture, enabling faster and better informed decision-making.

Methodology and challenges

The goal of the ML model presented in this study is to evaluate new logs and identify anomalous behavior. For that to be done, a clean dataset should be prepared, and the model needs to be configured with the right options to best perform its objectives. The model is then trained and tested, before being deployed to examine larger amounts of data.

Nearly every organization possesses unique datasets that can be leveraged to enhance and improve services. However, with this opportunity comes the critical responsibility of adhering to compliance standards and legal requirements. This responsibility is a key reason why we cannot disclose all the details of the training process and algorithms involved. Below we discuss the steps we have taken to achieve the goal, as well as the challenges you might encounter while training and applying similar models.

The dataset

A machine learning dataset is a collection of data used to train, validate, and test ML models. A dataset consist of various examples, each containing features (input variables) and, in supervised learning tasks, corresponding labels (output variables or targets).

Our dataset has been collected from, and is representative of, a large variety of previously investigated incidents. It describes advanced malware activities that we have observed in attacks by a variety of threat actors. More information about APT threat actors we track can be found here. The dataset contains a variety of indicators of malicious activities that were verified manually after being automatically collected to reduce the risk of inaccuracy.

A well-prepared dataset is foundational to the success of any ML project. The choice of dataset, its quality, and how it is prepared and split into training, validation, and test sets can significantly impact the model’s ability to learn and generalize new data. The adage of “garbage in, garbage out” applies here: if the dataset is flawed or poorly curated, the ML model’s predictions will also be unreliable.

Preprocessing

Preprocessing is a crucial step in a machine learning pipeline where raw data is transformed into a format suitable for training an ML model. This process involves cleaning the data, handling missing values, transforming variables into a scaled and normalized numerical representation, and ensuring that the data is in a consistent and standardized format. Effective preprocessing can significantly improve the performance and accuracy of an ML model.

One popular method we used for this transformation is term frequency–inverse document frequency (TF–IDF), which is a statistical measure used in natural language processing to evaluate the importance of a word in a document relative to a collection of documents. TF–IDF transforms raw text data into a set of machine-readable numerical features, which can be then fed to an ML model.

Implementation and training of the model

When dealing with text data, a common approach is to first transform the raw text into numerical features using techniques like TF–IDF and then apply an ML algorithm such as Random Forest to classify or analyze the data. TF–IDF is known to be efficient and versatile, while Random Forest is known for accuracy, reduced overfitting, and an ability to capture complex, non-linear relationships between features. The combination of TF–IDF with Random Forest allows handling high-dimensional data, while also providing robustness and scalability, very much needed to handle data with millions of entries daily.

A machine learning model reaches maturity when it performs consistently well on the kind of tasks it was designed for, meeting the performance criteria set during its development. Maturity is typically indicated by the following factors:

• Stable performance: The model exhibits consistent and satisfactory performance on both training and validation datasets.
• Generalization: The model generalizes well to new, unseen data, maintaining performance levels similar to those observed during testing.
• Diminishing returns from more training: After a certain point, adding more training data or fine-tuning the model further does not lead to significant improvements.

It is worth noting that model training does not necessarily end when it achieves maturity. To maintain model maturity, incremental learning is often needed, which means an ongoing process of updating and refining the machine learning model by incorporating new data over time. This approach is particularly important in dynamic fields where data distributions and patterns can shift, leading to the need for models that can keep up with these changes, which is exactly the case with the cybersecurity threat landscape.

During the development of our model, a target accuracy of 99% was set as a primary performance goal. Once the model achieved this level of accuracy, additional steps were implemented to further refine the output. These steps involved classifying the outcome in specific cases that required manual investigation.

Deployment and computational costs

When a model is ready, it can be integrated into a production environment where it can start making predictions on new data. This could be through an API, embedded in an application, or as part of a larger system. Continuous monitoring of the model’s performance to detect issues such as “concept drift” is essential to avoid degradation in accuracy, ensure reliable predictions, and maintain the model’s relevance in changing environments.

At the same time, both choices of TF–IDF and Random Forest can be computationally intensive. TF–IDF vectorization can lead to very large and sparse matrices. Random Forest models can become demanding when dealing with high-dimensional data and large datasets.

However, with the right capacities and hyperparameter fine-tuning, an optimal ML framework can be reached to permit the model to achieve its potential without sacrificing accuracy or efficiency. Example TF–IDF parameters that we found useful include frequency thresholds to allow, maximum features to extract, ngram range to match. Example Random Forest parameters that we found useful include the number of estimators, depth of the trees, split and leaf samples, and quality and impurity measurements.

Interpretability of results

In ML, and depending on the algorithm used, interpreting the model to understand how it makes predictions could be possible. This could involve analyzing the importance of features, visualizing decision trees, or using other evaluation tools.

However efficient, each model has its limitations that are important to know. Random Forests can be less interpretable than other models, especially with large numbers of input features. TF–IDF by design tends to give higher importance to rare items, which can sometimes be noisy or irrelevant in real-world applications. These feature limitations were acknowledged during development and considered in model fine-tuning.

The outcomes of machine learning during model training play a crucial role in guiding the development, refinement, and optimization of the model. These outcomes provide valuable feedback that helps data scientists and ML engineers make informed decisions to improve the model’s performance, guide adjustments, and ensure the final model is robust, generalizes well, and meets the desired criteria.

Findings: new cyberthreats discovered

As we started this study, we kept in mind that the usage of ML in log analysis enables the discovery of previously unknown cyberthreats by analyzing vast amounts of data and uncovering patterns. The model can process and learn from millions of data points in real time, pointing out subtle indicators that may signal the presence of a new or advanced threat. However, the results we got exceeded our expectations: the model revealed thousands of new advanced threats. As of H1 2024, ML findings represented 25% added detections of APT-related activities.

Here are a few examples of indicators of compromise found using ML in the past year. Once inspected, these indicators were quickly revealed to be part of malicious activity. The incidents involving these IoCs are not further investigated here, but our analysis of these and similar findings has been published (for example, here and here) or will be, in other posts on this site or in private threat intelligence reports.

Future directions

The future of using ML for telemetry analysis holds exciting possibilities, with several advancements on the horizon that could further enhance threat detection capabilities. One promising area is the integration of deep learning techniques, which can automatically extract and learn complex patterns from raw data. We already use deep learning in some of our products, and applying it to threat hunting could potentially further improve detection accuracy and uncover even more sophisticated threats.

Another area of exploration is reinforcement learning, where models can continuously adapt and improve by interacting with dynamic cybersecurity environments. This could lead to more proactive defense mechanisms that not only detect but also respond to threats in real time.

Additionally, federated learning presents a significant opportunity for collaborative threat detection across organizations while preserving data privacy. By allowing models to learn from decentralized data without sharing the actual data, federated learning could facilitate the creation of more robust and generalizable threat detection models.

Conclusion

The integration of ML into cybersecurity has already demonstrated its transformative potential by enabling the detection of novel cyberthreats that traditional methods might overlook. Through the analysis of vast and complex logs, ML models can identify subtle patterns and IoCs, providing organizations with a powerful tool to enhance their security posture. The examples of cyberthreats discovered over the past year underscore the efficacy of ML in uncovering threats across various industries, from government to finance.

These technologies will not only improve detection accuracy but also enable more proactive and collaborative defense strategies, allowing organizations to stay ahead of the ever-evolving cyberthreat landscape.

In this post, we have evaluated the utilization of ML models on our KSN global threat data, which has led us to reveal thousands of new advanced threats. The journey of refining ML models through meticulous dataset preparation, preprocessing, and model implementation has highlighted the importance of leveraging these technologies to build robust, adaptable, and scalable solutions.

As we continue to explore and enhance these capabilities, the potential for machine learning to reshape cybersecurity and protect against increasingly sophisticated threats becomes ever more apparent. The future of cybersecurity lies in our ability to ethically harness these tools effectively, ensuring a safer digital environment for all.

securelist.com/machine-learnin…

L’unica sicurezza che il decreto 1660 implementa è la sicurezza dei governi liberticidi di fare quello che vogliono rimanendo indisturbati, impunemente al riparo da qualunque protesta.

FIRMARE per FERMARE il DDL1660:

slowforward.net/2024/10/01/fer…

#ddl1660 #DDL1660 #ddl #DdlSicurezza1660 #neofascismo #statodipolizia

@Storia
@Storiaweb

La vicenda umana e professionale di Carlo Vicari, l’ultimo soldato pontificio che dismise la gloriosa uniforme della Truppa di Finanza Pontificia, il Corpo che vantava, già allora, il prestigio di essere stata una delle più remote Milizie Doganali sorte in Italia

giornidistoria.net/carlo-vicar…

#Europol ha recentemente ospitato l'annuale European Police Chiefs Convention (#EPCC) all'Aia, dove sono state discusse le principali sfide che la polizia moderna deve affrontare. L'evento, co-ospitato dalla presidenza ungherese del Consiglio dell'Unione europea, funge da piattaforma leader per la cooperazione internazionale di polizia, facilitando gli incontri tra alti funzionari di polizia degli Stati membri dell' #UE e di tutto il mondo per affrontare questioni operativamente rilevanti attraverso discussioni bilaterali e sessioni plenarie. L'EPCC ha riunito oltre 450 delegati da 50 paesi, concentrandosi su strategie collaborative per combattere le minacce criminali emergenti e migliorare la cooperazione internazionale. Tra i relatori principali Laura Kövesi, Procuratore capo europeo (foto), János Balogh, Alto commissario della polizia nazionale ungherese, Drew Harris, Commissario An Garda Siochana, Hans Leijtens, Direttore esecutivo di Frontex, e Ivan Vyhivskyi, Capo della polizia nazionale dell'Ucraina.

L'EPCC ha inoltre evidenziato la necessità di un accesso legale ai dati e di modernizzare gli strumenti delle forze dell'ordine per accedere alle informazioni digitali.
Dalle discussioni sono emersi temi comuni, tra cui la necessità critica di concentrarsi su obiettivi di alto valore, la prevenzione come componente chiave nella lotta contro la criminalità organizzata, le aziende tecnologiche, i fornitori di servizi e le piattaforme di intelligenza artificiale, le solide partnership con entità private e l'evoluzione delle minacce ibride. Europol ha inoltre presentato gli Europol Excellence Awards in Innovation 2024 per evidenziare i progetti più innovativi nelle forze dell'ordine, premiate per il loro lavoro Croazia, Estonia e Paesi Bassi .