Sana - the world´s first artificially intelligent teacher.
|Coordinator||Sana Labs AB|
|Funding from Vinnova||SEK 300 000|
|Project duration||March 2017 - July 2017|
Purpose and goal
The aim of the project was to perform research/analysis of data sets and algorithms for the creation a scientific foundation for the following experimental development, training and evaluation of selected algorithms with the ultimate goal to find a relevant and feasible algorithm to be included in the Brain. The project was conducted successfully and all the goals were satisfactory fulfilled.
Expected results and effects
In the project, we have created a new data model forming the Brain algorithm that has been trained and evaluated on public data sets. Throughout the project, we have proven that our model provides better predictions for the recommendation system for the optimal teaching method for each student than the state-of-art models. The results of this project will serve a basis for the following developmental efforts of the Brain and Sana technology.
Planned approach and implementation
During the course of the project, we have conducted industrial research on algorithms and mathematics, made research and analysis of relevant datasets, created and implemented algorithms via a production ready Deep Learning framework, trained and evaluated the algorithms performance, and compared with the state-of-the-art research. We have managed to develop and validate a (simple) algorithm that will form the basis of the continuous development of the Brain functionality.