MyDataModels Offers Free Access To Its Self-Service Machine Learning Platform For Public Organizations Fighting COVID-19.
As you read these words, can anyone honestly say that COVID-19 has not, does not or will not impact our health or the health of the people we care for? We, at MyDataModels, can’t make these claims as some of us have already been touched by the virus. It was difficult to realize that an invisible, intangible Tsunami just hit us. The world around us just stopped as we all went into containment. And we dreamt for a minute that we could turn into modern time Alexander Flemings finding the new penicillins that would save lives…
Yet, we are not medical doctors, nor biologists, nor epidemiologists. We’re plain, passionate, genuine data scientists. That’s who we are. That’s what we do. And we want to help. How?
The global COVID-19 pandemic is a rare unexpected event, a typical ‘Black Swan’* You do not have millions of data related to this disease yet and you are running out of time. When most of today’s machine learning systems need to be trained on lots of historical data, ours can give interesting results with few in a matter of minutes. And when the present suddenly stops looking like the recent past, we can help.
We realized that some of our customers in the medical world are not data scientists and succeed to get important health related results with our platform. For instance, is there a correlation between cells contained in a blood sample and people suffering from lung cancer? Yes there is and our customers (medical doctors with no data science knowledge in this case) can provide a fairly good estimate of whether a patient is at risk of lung cancer from a simple blood sample using our solution.
Most COVID-19 related issues which can be tackled by artificial intelligence are similar to this one. There is a small amount of historical data available (instead of millions, thousands). The people who own and understand the data are not data scientists. They are field experts, epidemiologists, medical doctors, researchers.
They ask themselves questions such as: can we identify quickly which patients are at risk of developing a severe form of the illness, i.e. not all patients with diabetes are at risk equally? Can we combine data about blood type, gender, initial viral load, overexposure to the virus in order to make a precise risk estimation? Which patients should be protected, isolated and cared for in priority in order to avoid the urgent care unit? Can we identify several phases in the disease and provide ad hoc treatment? Can we provide self diagnosis questionnaires which go beyond a simple decision tree and take into account subtle differences learned from previous cases?
That’s when we realized that even though we are no Alexander Fleming, we can help the Alexander Flemings of this world get quicker to the new penicillins. So that’s why we have decided as a team to offer free access to our solution to researchers in public institutions who have some data but no easy and simple way to analyse them in depth.
Our solution is a simple-to-use data science software. Because time is of the essence. That’s why our senior data scientist team has also volunteered their time and expertise to support researchers who want some help in jump starting their data science project. They will help you on their own private time, free of charge.
Let find the “small signals” buried in your data!