Machine learning and automatic reasoning
Listen to podcast highlights: How AI Reasons and Learns
Although machine learning and machine reasoning are two powerful AI technologies, they have two different approaches that solve different kinds of problems. In mechanical reasoning, we are talking about human common sense, where ideas and concepts are represented as symbols in a computer system. Then logic or rules are used to combine these symbols to reason your way to an outcome. Machine learning is different. It is best described as the use of advanced statistical techniques to find subtle patterns in very large collections of data.
The first approaches to AI were in the areas of reasoning. Logical reasoning has addressed problems related to different types of games, such as playing chess, i.e. problems that tend to have all the information and options at your fingertips, and where you reason until a specific result from certain initial conditions. This approach was popular for the first three or four decades of the field. But people quickly discovered that there were certain problems that did not lend themselves to this type of approach.
How to make sense of big data sets, analyze the content of an image, or understand human speech are abilities where machine learning has succeeded. Most of what you hear about today called AI is some type or form of machine learning, which involves finding statistical patterns in very large collections of data so that we can classify new types of results .
Let’s take a relatable example of the two models in action. Imagine you are in a self-driving car or self-driving vehicle. When traveling from point A to point B, the car has to find or plan a route – this is usually done with reasoning. There is a simple set of prerequisites; there is a set of logical connections. Decisions have to be made about where the car should turn and how best to get from your home to your municipal office, for example. This route planning is a clear example of automatic reasoning. On the other hand, when your car uses its cameras to determine that it’s at a crossroads and needs to stop, that’s an example of machine learning. Thus, these two technologies combine and collaborate to allow us to solve certain types of problems.
|Machine learning can process large volumes of data and capture the hidden patterns needed to effectively predict outcomes. It tackles a predetermined problem, with clear inputs and expected outputs.||Automatic reasoning can be seen as an attempt to implement abstract thinking as a computational system and apply human common sense to analyze and translate vast knowledge and learned network data into clear, explainable information. It does this by providing more contextual knowledge, concepts, and rules that systems can obey and from which they can begin to create a model of the world around them.|
ML and MR in the telecommunications industry
For quite some time, machine learning models have been implemented in many different places. But over time, the complexity of the systems in which they have been implemented has increased. This is the case in telecommunications networks, where operations are becoming increasingly intelligent and automated.
With machiAs learning becomes industrialized through telecommunications networks, we are beginning to see some shortcomings of this technology:
- Machine learning relies on large amounts of learned data to make recommendations. Another solution is therefore necessary when there is less historical data to rely on.
- Machine learning models do not offer a simple way to trace the reasons behind recommendations.
- It can be difficult to consolidate and prioritize different advice from distinct machine learning agents.
This is where machine reasoning can complement machine learning.
Automatic reasoning solves problems by applying human common sense to learned data. It builds on the possibilities offered by machine learning, analyzing vast sources of knowledge and data to offer clear and explainable insight into the increasingly complex world of network operations, and finally aim intent-based networks. Mechanical reasoning is capable capture business intent and break it down into achievable network goals and KPIs. It can then autonomously balance and prioritize them based on defined business intent to make recommendations and decisions that enable further automation of network operations.
In addition to tackling increased complexity, we are also seeing an increase in end-user expectations on the network to a level never seen before. For example, gamer expectations have never been higher when it comes to latency and throughput. It’s also a huge market, so you really have to meet expectations. More and more people are turning to networks to provide instant, smooth, lag-free connectivity – essentially, great user experiences. This is the reality that service providers need to be prepared for. This is also where machine learning and machine reasoning can have significant benefits. Independently, machine learning is a great solution for solving a single problem, but it’s less suited to tasks that require more careful, deliberate, and explicit thinking. Generalizing or dealing with problems that are different from the original task is difficult for machine learning. However, machine reasoning adds these much-needed skills to machine learning, contributing to more abstract thinking and giving machines the power to make new connections between facts, observations, and the various things they can already be trained. to do with machine learning.
The Challenges and Prospects of Machine Learning and Machine Reasoning
The immediate challenge with the expansion of AI into networks is that with the growth of AI, products, systems and applications must reflect human values. Consideration must also be given to how the design – and the transparency, or lack thereof – of these systems might codify values or promote or obscure certain human viewpoints. Machine reasoning is crucial here in addition to machine learning, as it provides recommendations that are explainable. This allows humans to trace all decisions made throughout the process, which increases the verifiability and explainability of the system. However, to ensure the implementation of fully unbiased AI systems, laws and technical approaches must also be implemented industry-wide to help identify when this is happening and eliminate it. or ban it as we continue to advance with this important technology.
The telecommunications industry is built on trust and maintaining that trust is crucial as we move into a post-Covid world. For example, how we adapted and used technology to work from home during the pandemic was something we didn’t expect to see for another three years, which sets the bar high.
As new requirements and technologies come into play, we will also see the expansion of technology platforms, limitless connectivity and a digitally driven society. There is a lot to come, so the focus is on organizations such as the European Commission and guidelines for Trustworthy AI. We need to ensure that we embed secure and trusted AI into our processes and management systems as we further implement AI technologies in the telecommunications industry.
We have an exciting future ahead of us. Machine learning and machine reasoning are a critical part of building the future of autonomous networks, trusted AI systems, and platforms, through which customer expectations, use cases, and connectivity will be executed. AI-based solutions are key to making this evolution happen.
Listen to the full podcast episode, How Does AI Reason and Learn?
Read our previous blog post in this series:
Can we use AI to build our future society?
What is the relationship between AI and 5G?
Want to know more?
Read our previous blog post in this series, What is the relationship between AI and 5G?
Learn more AI Bias and Human Rights: Why Ethical AI Matters.
Learn more about Explainable AI and how humans can trust AI.
Learn more about 5G monetization through intent-based network operations