Training your model is hands down the most time consuming and expensive part of machine learning. Training your model on a GPU can give you speed gains close to 40x, taking 2 days and turning it into a few hours. However, this normally comes at a cost to your wallet.
Source: hackernoon.com
Google’s new Cloud AutoML Vision is a new machine learning service from Google Cloud that aims to make state of the art machine learning techniques accessible to non-machine learning experts. In this post I will show you how I was able, in just a few hours, to create a custom image classifier that is able to distinguish between different types of poisonous Australian spiders. I didn’t have any data when I started and it only required a very basic understanding of machine learning related concepts.
Read More
From November 2017 to January 2018 the Google Brain team hosted a speech recognition challenge on Kaggle. The goal of this challenge was to write a program that can correctly identify one of 10 words being spoken in a one-second long audio file. Having just made up my mind to start seriously studying data science with the goal of turning a new corner in my career, I decided to tackle this as my first serious kaggle challenge.
Read More
In this post, we’ll implement a deep neural network with one hidden layer (and ReLu and softmax activation functions) purely in SQL. The end-to-end steps for neural network training including the forward pass and back-propagation will be implemented as a single SQL query on BigQuery. As it runs on Bigquery, in effect we are performing distributed neural network training on 100s to 1000s of servers.
Read More