Application of deep learning to predict and recommend consumer online purchase behaviour at a …

Preferred Disciplines: Mathematician – Computer scientist (Master, Post-Doc) studying or with studies in Deep LearningCompany: N/AProject Length: 4 monthsDesired start date: February 2016Location: Montreal, QCNo. of Positions: 2Preferences: Ability to have the interns on-site is the only requirement. About the Company:  N/A Project Description: The purpose of this project is to develop a recommender system able to recommend products cross-databases analysing user behavioral data and consolidating databases through image classification and NLP  in order to identify equivalent products in very large product spaces. Research Objectives: Classification of products. Convolutional neural networks for image classification: Train from a random initial. Transfer knowledge from a trained model (fine-tuning) Organization in the image – identify product positions in each image. Comparison between products. Find equivalent products. Product classification in extremely large label spaces. …


Link to Full Article: Application of deep learning to predict and recommend consumer online purchase behaviour at a …