Visit site

RecomendacionLiterarIA

Made by@luisehica avatar@luisehica
RecomendacionLiterarIA is a machine learning project using supervised learning techniques to predict book ratings. In this initial version, several models were tested, but all exhibited low performance. This could be due to limitations of the available data. Additionally, the high number of categories and authors resulted in over 600 features, likely introducing significant noise.RecomendacionLiterarIA screenshot

Latest updates

9 January 2025
@luisehica avatar
Lui Higuera• Complex System Researcher Wanabe - AI Builder

I improved my features by using PCA to reduce dimensionality from one-hot encoded categories (300+) and word embeddings for authors. Testing with random forest, our best model, only improved R² by 0.03. Evidently, these features cannot effectively predict book ratings.#RecomendacionLiterarIA