Tech Bytes - Daily Digest: Daily Tech Digest

The engines of AI: Machine learning algorithms explained

Machine learning algorithms train on data to find the best set of weights for each independent variable that affects the predicted value or class. The algorithms themselves have variables, called hyperparameters. They’re called hyperparameters, as opposed to parameters, because they control the operation of the algorithm rather than the weights being determined. The most important hyperparameter is often the learning rate, which determines the step size used when finding the next set of weights to try when optimizing. If the learning rate is too high, the gradient descent may quickly converge on a plateau or suboptimal point. If the learning rate is too low, the gradient descent may stall and never completely converge. Many other common hyperparameters depend on the algorithms used. Most algorithms have stopping parameters, such as the maximum number of epochs, or the maximum time to run, or the minimum improvement from epoch to epoch. Specific algorithms have hyperparameters that control the shape of their search.

How to Build a Cyber-Resilient Company From Day One

Despite your best proactive measures, some cyber threats will infiltrate your defenses. Reactive defenses, such as firewalls and antivirus software, help to minimize damage when these incidents occur. Firewalls monitor and control incoming and outgoing network traffic based on predetermined security rules, forming the first line of defense against cyber threats. Antivirus software complements firewalls by detecting, preventing and removing malicious software. Intrusion Detection and Prevention Systems (IDS/IPS) monitor your network for suspicious activities and potential threats, alerting you to a potential attack and, in some cases, taking action to mitigate the threat. Encryption is another valuable reactive measure that involves making your sensitive data unreadable to anyone without the appropriate decryption key, thus protecting it even if it falls into the wrong hands. Security Information and Event Management (SIEM) systems provide real-time analysis and reporting of security alerts generated by applications and network hardware. They help detect incidents early and respond promptly.

Quantum Algorithms vs. Quantum-Inspired Algorithms

Quantum-inspired algorithms refer usually to either of the two: (i) classical algorithms based on linear algebra methods — often methods known as tensor networks — that were developed in the recent past, or (ii) methods that attempt to use a classical computer to simulate the behavior of a quantum computer, thus making the classical machine operate algorithms that benefit from the laws of quantum mechanics that benefit real quantum computers. On (i), while the physics community has leveraged these methods to address problems in quantum mechanics since the 70s [Penrose], tensor networks have an independent origin as far back as the 80s in neuroscience as well, as there is nothing really quantum behind them; it really is just linear algebra. For (ii), the process of emulating a quantum system falls back on the limitations of classical hardware. It is very hard to emulate classically the full dynamics of a large quantum system for the exact same reasons that one wants to actually build a real one! So, does this mean that quantum-inspired algorithms are bogus? Not really.

Operator survey: 5G services require network automation

"Private 5G" and "network slicing" rank second and third, respectively. Heavy Reading expects their importance and popularity to increase as additional operators deploy 5G SA and can support full autonomy. "Performance SLAs for enterprise services" is currently the lowest ranking (fifth) of all service choices but is likely to be a valuable market, especially for network slicing and private 5G. "Connected devices (e.g., cars, watches, other IoT devices)" ranks just above performance SLAs in fourth. Internet of Things (IoT) is a sizeable market within 4G, but the massive machine-type communications (mMTC) use case has yet to be realized in 5G, as technologies such as RedCap remain underdeveloped. Smaller operators have a different opinion from larger operators on the revenue growth question. For mobile operators with less than 9 million subscribers, private 5G ranks first. This result perhaps indicates that smaller operators feel they are already exploiting eMBB services and see little scope for further revenue growth with SA.

No cyber resilience without open source sustainability

Open source sustainability is a problem: maintainers of popular software projects are often overwhelmed by issues and pull requests to the point of burnout. Donations have emerged as one solution, and are regularly provided by governments, foundations, companies, and individuals. Yet, as excerpts of recent drafts of the CRA indicate, it could threaten to undermine sustainability by potentially introducing a burdensome compliance regime and potential penalties if a maintainer decides to accept donations. The result will be less resources flowing to already under resourced maintainers. Open source projects are often multi-stakeholder: they receive contributions from developers building as individuals, volunteering in foundations, and working for companies, large and small. The current text would regulate open source projects unless they have “a fully decentralised development model.” Any project where a corporate employee has commit rights would need to comply with CRA obligations. This turns the win-wins of open source on its head. Projects may ban maintainers or even contributors from companies, and companies may ban their employees from contributing to open source at all.

Building Trust in a Trustless World: Decentralized Applications Unveiled

In a DApp, smart contracts are used to store the program code and state of the application. They replace the traditional server-side component in a regular application. However, there are some important differences to consider. Computation in smart contracts can be costly, so it's crucial to keep it minimal. It's also essential to identify which parts of the application require a trusted and decentralized execution platform. With Ethereum smart contracts, you can create architectures where multiple smart contracts interact with each other, exchanging data and updating their own variables. The complexity is limited only by the block gas limit. Once you deploy your smart contract, other developers may use your business logic in the future. There are two key considerations when designing smart contract architecture. First, once a smart contract is deployed, its code cannot be changed, except for complete removal if programmed with a specific opcode. It can be deleted if it is programmed with an accessible SELFDESTRUCT opcode, but other than complete removal, the code cannot be changed in any way.

How the upcoming Cyber Resilience Act will impact privacy

The Cyber Resilience Act has several positive implications for privacy. Firstly, by enforcing strict standards of cybersecurity in the development and production of new devices, the Act creates an ecosystem where security is ingrained in the product development cycle. Secondly, by creating the reporting obligations, the Act ensures that vulnerabilities are addressed promptly, reducing the risk of personal data breaches and protecting the privacy of individuals. Third, the Act empowers consumers by ensuring they are informed about the vulnerabilities in their devices and the measures they can take to protect their personal data. From the perspective of data controllers, particularly those who serve as manufacturers of devices regulated by the Act, compliance requirements are raised to an even higher threshold. ... Additionally, they will have to comply with reporting obligations regarding vulnerabilities, even those that have already been fixed, regardless of whether personal data was affected or not. Neglecting to fix known vulnerabilities may also result in reputational consequences for data controllers.

Crafting a cybersecurity resilience strategy: A comprehensive IT roadmap

In recent years, there has been a significant increase in the demand for cybersecurity professionals due to the growing importance of protecting sensitive information and systems from cyber threats. Organizations are allocating larger budgets to enhance their cybersecurity measures, resulting in a surge in the number of job opportunities in this field. According to the latest Cyber Security Report by Michael Page, companies are actively seeking skilled cybersecurity talent to address their security challenges. The report reveals that globally, more than 3.5 million cybersecurity jobs are expected to remain unfilled in 2023 due to a shortage of qualified professionals. This shortage has created a sense of desperation among companies, as they struggle to find suitable candidates to fill these critical roles. India is projected to have over 1.5 million vacant cybersecurity positions by 2025, underscoring the immense potential for career growth in this field. To effectively address the ever-changing risks of digitalization and increasing cyberthreats, it is crucial for organizations to implement a continuous security program.

The rise of OT cybersecurity threats

There is a need for a separate security program for OT that includes different tools, governance, and processes. Companies can’t simply extend their IT security program to OT, as the differences between the two domains are too great. It may require two security operation centers (SOCs), which adds to the complexity and costs of cybersecurity management. Bellack explains that some CEOs or CIOs underestimate the risks associated with an OT attack. “It’s a relatively new set of risks and a lot of executives don’t understand that they are indeed in danger,” Bellack says. “Companies build smarter, faster, cheaper factories using digital technologies because it’s great for business. But it also expands their attack surface, and many people in charge don’t realize the impacts or what they need to do to protect themselves.” ... “Machines are components in a complex, revenue producing infrastructure that is a mix of physical, digital, and human elements. Safety and availability are the key focus, and security is sometimes forced to take a back seat if either of those may be compromised,” explains Boals.

Quote for the day:

"Practice isn't the thing you do once you're good. It's the thing you do that makes you good." -- Malcolm Gladwell

Tech Bytes - Daily Digest

Pages

Daily Tech Digest - July 16, 2023