|
The next step is neuro-symbolic computation.
We are the world leaders in this area. Our approach combines the power of logical inference and statistical machine learning to tackle health challenges with model-driven approaches.
Our neuro-symbolic computation approach is a type of responsible A.I. that does not rely on black-boxes to get and understand answers to critical questions in areas such as medicine.
|
This video produced by the journal Nature introduces our approach to
Causal Diagnostics using Causal Deconvolution
Causal Diagnostics |
|
Despite the many advances in medical technology, practical clinical diagnosis is still largely based on symptom pattern-matching. OIA is changing this by leveraging our years of research in some of the best universities in the world developing the area that we call causal diagnostics, a cutting-edge and responsible model-based approach to medicine different from other AI approaches such as those proposed by statistical machine learning and deep learning that are unable to provide explanations for their many arbitrary choices confounding all factors in favour of min-maximising random parameters at all costs in favour of classification tasks.
To change this, at OIA we have created and continue creating a family of novel and interdependent solutions for rapid and affordable generation of personal longitudinal data using sophisticated algorithms and repurposing off-the-shelf electronics augmented with responsible AI algorithms in order to achieve the production and commercialisation of the first portable and personal general-purpose medical machine. |
Longitudinal, non-invasive, (quasi) real-time and cost-effective health monitoring is hugely important if we are to improve public health and deliver the promise of personalised medicine for super early accurate diagnosis.
Causal diagnostics will revolutionise medicine by refocusing on first principles, generating mechanisms and root causes to redesign drugs and find therapies for diseases beyond dealing with symptoms only; no more 'no cure' or 'too late' arguments. The training and inference engines behind this type of algorithmic causal diagnosis are based on what is identified as neuro-symbolic computation and relying on both mind and machine to fight disease. |
Neuro-symbolic Computation
|
|
|
While techniques such as machine & deep learning (e.g. deep neural networks) are very powerful in tasks such as classification, they are only so under very controlled conditions and are ill-equipped to deal with some basic cognitive functions that we as humans take for granted, including basic logical inference and hypothesis modelling.
For example, no amount of big data or sophisticated statistics can make a neural network to e.g. perform a basic arithmetic operation if it has not seen it before. The combination of symbolic computation and statistical machine learning is what we have been successfully doing for years and the time is ripe to use these methods in combination with machine & deep learning in business to solve challenges. |
|
AI as a service &
our AI cloud
|
|
|
Most A.I. instils lofty expectations but fails to effectively link A.I. solutions to business goals, or else is too detailed and technical in nature and misses the big picture. We can help make sure solutions are aligned with your organisation goals and expectations.
While our science develops responsible machine learning tools, it may quite difficult to integrate them into existing processes simply because it takes time for companies and institutions to catch up. Often companies do not have data in the required form and the data scientists wait idle for months. Often nobody knows how to integrate data science into existing structures and many other problems. We help your executive team at every stage of the adoption of a responsible A.I. solution, from conception to prototyping to implementation, and our focus is always on the point of intersection of a client’s goals and potential technological solutions. We don't let A.I. 'libraries' dictate what we can or cannot do, thus sparing you the experience of consigning your data to a black hole, with no return. Some of our B.I. customers |
We bring the experience of a leading team to help make the best of a solid AI strategy that we can design for you
|
|
|
Personalised Medicine
Selection of our publications
We have shown how our algorithms can deal with complex biological systems even in the face of very little data by using model-based approaches combined with sophisticated data meaning extraction.
