During my Pythonic data science team coaching I see various problems coming up that I’ve helped solve before. Based on these observations and my prior IP design and delivery for clients over the years I’ve put together a 1 day public course aimed at data scientists (any level) who want to be more confident with lower-risk approaches to delivering data science projects.

A simplified explanation for Understanding the Mathematics of Deep Learning

At: LoyaltyOne Location: Westborough, MAWeb: www.loyalty.comPosition: Consultant Category Manager / Analyst, Client Services

At: Vanguard Location: Malvern, PAWeb: www.vanguard.comPosition: Senior AI Architect

Compositional GAN Generative Adversarial Networks (GANs) can produce images of surprising complexity and realism, but are generally modeled to sample from a single latent source ignoring the explicit spatial interaction between multiple entities that could be present in a scene. Capturing such complex interactions between different objects in the world, including their relative scaling, spatial layout, occlusion, or viewpoint transformation is a challenging problem. In this work, we propose to model object composition in a GAN framework as a self-consistent composition-decomposition network. Our model is conditioned on the object images from their marginal distributions to generate a realistic image from their joint distribution by explicitly learning the possible interactions. We evaluate our model through qualitative experiments and user evaluations in both the scenarios when either paired or unpaired examples for the individual object images and the joint scenes are given during training. Our results reveal that the learned model captures potential interactions between the two object domains given as input to output new instances of composed scene at test time in a reasonable fashion. …

We’ve heard from customers that scaling TensorFlow training jobs to multiple nodes and GPUs successfully is hard. TensorFlow has distributed training built-in, but it can be difficult to use. Recently, we made optimizations to TensorFlow and Horovod to help AWS customers scale TensorFlow training jobs to multiple nodes and GPUs. With these improvements, any AWS customer can use an AWS Deep Learning AMI to train ResNet-50 on ImageNet in just under 15 minutes.

Ho, ho, ho! It’s almost Christmas time and I don’t know about you, but I can’t wait for it! And what can be a better way of killing the waiting time (advent!) than participating in excellent Advent od Code. Big thanks to Colin Fay for telling me about it! It’s a series of coding riddles, one published every day between 1st and 25th of December. The riddles increase in difficulty level over time and they can be solved in any programming language, including R. After you’ve solved the first riddle of the day, the second one – a more difficult one – will be unlocked.

Unbiased Implicit Variational Inference (UIVI) We develop unbiased implicit variational inference (UIVI), a method that expands the applicability of variational inference by defining an expressive variational family. UIVI considers an implicit variational distribution obtained in a hierarchical manner using a simple reparameterizable distribution whose variational parameters are defined by arbitrarily flexible deep neural networks. Unlike previous works, UIVI directly optimizes the evidence lower bound (ELBO) rather than an approximation to the ELBO. We demonstrate UIVI on several models, including Bayesian multinomial logistic regression and variational autoencoders, and show that UIVI achieves both tighter ELBO and better predictive performance than existing approaches at a similar computational cost. …

By Diogo Menezes Borges.