Hi! New AI Weekly is here! Enjoy your weekend reading AI news and don’t forget to share it with your friends 😉
Project Reports and Posters, Spring 2018 CS230: Deep Learning
How to build a Teachable Machine with TensorFlow.js
Neural Machine Translation with Attention – This notebook trains a sequence to sequence (seq2seq) model for Spanish to English translation using tf.keras and eager execution. This is an advanced example that assumes some knowledge of sequence to sequence models.
Papers with Code
Face recognition with OpenCV, Python, and deep learning
Scalable and accurate deep learning with electronic health records – Predictive modeling with electronic health record (EHR) data is anticipated to drive personalized medicine and improve healthcare quality. Constructing predictive statistical models typically requires extraction of curated predictor variables from normalized EHR data, a labor-intensive process that discards the vast majority of information in each patient’s record. Authors propose a representation of patients’ entire raw EHR records based on the Fast Healthcare Interoperability Resources (FHIR) format. They demonstrate that deep learning methods using this representation are capable of accurately predicting multiple medical events from multiple centers without site-specific data harmonization.
important observation: “regularized logistic regression essentially performs just as well as Deep Nets” (http://bit.ly/2MTST9X)
RUDDER: Return Decomposition for Delayed Rewards – Authors propose a novel reinforcement learning approach for finite Markov decision processes (MDPs) with delayed rewards. In this work, biases of temporal difference (TD) estimates are proved to be corrected only exponentially slowly in the number of delay steps. Furthermore, variances of Monte Carlo (MC) estimates are proved to increase the variance of other estimates, the number of which can exponentially grow in the number of delay steps.
Retro Contest: Results – The first run of Retro Contest — exploring the development of algorithms that can generalize from previous experience — is now complete. Though many approaches were tried, top results all came from tuning or extending existing algorithms such as PPO and Rainbow.