Machine learning from the point of view of Similarity Search. Why? Because if you tell a physician that a patient will have XYZ disease in 10 years the industry will make the prediction and maybe they will bring a confidence result. But that does not say much about the justification of the prediction. Similarity brings a very powerful and natural justification. For example our system could make a prediction with the following description: the patient will develop disease XYZ because this patient is similar to K patients in the following diagnosis: “icdn1, icdn2, icdn3” and the following similar family history and here are the phone numbers of the physicians who took care of the K patients.
Also, the industry solves problems by using Euclidean vectors, there is nothing wrong about this but if you use metric and non-metric spaces then you can model data with more freedom. For example: multi-variate time series, trees graphs and any kind of data shape....
Machine learning from the point of view of Similarity Search. Why? Because if you tell a physician that a patient will have XYZ disease in 10 years the industry will make the prediction and maybe they will bring a confidence result. But that does not say much about the justification of the prediction. Similarity brings a very powerful and natural justification. For example our system could make a prediction with the following description: the patient will develop disease XYZ because this patient is similar to K patients in the following diagnosis: “icdn1, icdn2, icdn3” and the following similar family history and here are the phone numbers of the physicians who took care of the K patients.
Also, the industry solves problems by using Euclidean vectors, there is nothing wrong about this but if you use metric and non-metric spaces then you can model data with more freedom. For example: multi-variate time series, trees graphs and any kind of data shape.
More information

Recommendations

Sonal Mane

Arnoldo has been instrumental in launching simMachines and we've been working on various MSFT opportunities since 2013. At the time, Arnoldo was prototyping the solution and working on identifying product market fit. Over the past few years, Arnoldo has expanded base from T-rex in St. Louis to 1871 in Chicago and proven that the technology behind simMachines is solid and has run several pilots in the industry across political campaigns and other prediction projects that have yielded positive results. I highly recommend this team as they have a strong combination of technology and business skills between Robert and Arnoldo. I'm looking forward to continuing work with this amazing company as build national startup-focused initiatives at MSFT.

Martin Schray

Arnoldo and Robert make the perfect leadership team. Arnoldo has such deep knowledge and passion for data science and his years of research have yielded an unparalleled approach to making similarity searches fast (nearly immediate) and with very little computing power (that is breakthrough innovation). Robert has built, managed and sold several tech startups. Robert knows business, but he also knows how to motivate and manage technical teams to develop offerings that the market needs and investors want. SimMchines technology offerings can quickly solve time series based multivariate problems that exist in financial services and health tech/biotech verticals in ways that create clarity and fidelity for business and scientific decision makers. Their offerings could easily live behind Azure API management and Azure Marketplace and provide real benefits to Microsoft and our customers.

Adarsh Arora Founder and CEO at Reputada

This is potentially a ground breaking technology in data analytics because of its ability to provide "back tracing" of how a result was obtained. I would highly recommend this company to be part of your program.

Steven Gould

simMachines has a very powerful data analytics platform that offers significant advantages in terms of speed and accuracy. The tool can be used in a large variety of situations such as financial services, trading, data security, many application in the healthcare/bio space and multiple others as well. Arnoldo is a very talented and skilled data scientist and bob is a ver savvy and experienced serial entrepreneur. I am extremely impressed with both their technology and their combined management talent and expertise. I believe they would be an excellent addition to your program.