The new algorithm is based on an improved version of the popular BPR (Bayesian Personalized Ranking), which is now one of the world standards. The new product also bypasses Netflix’s Mult-VAE algorithm, providing 10% more accurate recommendations. It took more than 200,000 GPU hours and 15,000 experiments to achieve these results.
The essence of the work is that existing algorithms for recommendation systems often have many versions that differ in efficiency. T-Bank Artificial Intelligence Research scientists created a more effective solution by examining and improving the components of the BPR algorithm. The new model outperformed existing options in terms of accuracy by almost 50%, making it attractive for use in e-commerce, streaming services and education platforms.
Recommender systems researcher Alexander Milogradsky said the new model shows how the correct implementation of algorithms can outperform more modern versions and highlights the importance of a detailed approach to development.
The discovery of Russian scientists was presented at the ACM RecSys conference in Bari, Italy. Attendance at the event was limited and only 17% of applications were selected.
Source: Ferra
I am a professional journalist and content creator with extensive experience writing for news websites. I currently work as an author at Gadget Onus, where I specialize in covering hot news topics. My written pieces have been published on some of the biggest media outlets around the world, including The Guardian and BBC News.
