Dyna
recsys
BENEFITS
Maximize the trustworthiness and the accuracy of recommendations with adaptative ranking and weights
With a robust, reliable, explainable and sustainable learning model
Using pre-processing data, trained feature sets and dynamic learning of weights to lower the impact of unreliable data.
Key points
Maximizing the quality and the reliability of recommendations, with continuous learning of weights and adaptative ranking to better fit users requirements, with less computationnal ressources
Dynamic
Weight of sources are dynamically updated according to induced accuracy and recall.
Ranking calculations are iterative and adapted to improve robustness
Sustainable
Based on predefined lenght, pre processed data and pre trained weight to avoid excessive resources computation
Smart
Using lean machine learning to improve data sources quality, to reduce noises and to avoid corrupted data
