Hi,

Thank you for the clarification.  Our single-blinded study conducted in 2016 had a primary objective to discern the difference between those who have love from those who do not.  Once we were able to define the variables that had a significant affect on love, we then designed the principles to create our application.

The app in itself was not reintroduced to the same subject group studied previously.  The app will apply machine learning based on facial recognition (vision technology) to determine the likelihood of physical attraction.  The continued iteration of user selection will improve our data models which in turn improves our deep learning process of matching users.  This convolutional neural network that is applied resembles the neural network translation found on our website.  Our website is able to translate text in over 100 languages using similar technology.

The subsequent features utilized by users (commits) will then apply separate data models to determine the likelihood of two individuals starting a relationship and the length of duration.  In the world of machine learning, the training data is applied over and over and the "learner" gets better with every iteration.  We reduce the chance of false input by applying an a/b testing scheme. 

So to answer your question in summary:

The study was used to create the design of the app.  The app will then be tested towards the end of the year to improve its effectiveness.  Let me know if I answered your question.

Hi,

I don't know if I exactly understand the question.  When you say "test those same acquaintances without him and the whole system", are you referring to our study and the use of the app?