In early January I was reached out to by the nyc blockchain group asking if I could teach whatever I know about Solidity. I said yes. The audience was undergrads completeing coursework in project management and computer science. Not all of them knew what bitcoin was so I covered a lot of ground in a short amount of time.
There are lots of materials on how to train neural networks but not as many showing how to deploy them to power real services. In this post I’ll share a minimal example of how I train tensorflow models and deploy them on Sagemaker. I’ll cover items that I found intimidating so hopefully they become less intimidating to you.
An app that’s tailored exclusively to English-speaking users won’t compete in the overall market. You could make 10x the revenue you currently make by adding global localization to your app. You limit your own chances of reaching a larger customer base if you don’t localize your business for users who speak different languages and live in different regions.
Running a full bitcoin node on osx is a good way to familiarize yourself with blockchain administration. You can explore the ledger locally and you don’t suffer any of the risks Simple Payment Verification nodes suffer from. You can “Be your own Bank” or so they say. If you have ~300 GB of free disk space you should try it out.
Gradient boosting decision trees is the state of the art for structured data problems. Two modern algorithms that make gradient boosted tree models are XGBoost and LightGBM. In this article I’ll summarize each introductory paper.
CryptoZombies is one of the best free beginner Solidity courses for learning Ethereum Smart Contract development. In this post I review the 6 lessons of the curriculum. I hope this feedback helps improve the site and motivates others to go through it as well. UPDATE: they updated cryptozombies
Big data analytics tools are a constantly shifting landscape. Most organizations want to be data driven, this means collecting your data into a data warehouse where it can be explored. The two biggest solutions in the space are AWS Redshift and Google BigQuery. Here is what your big data administrator needs to know to choose a solution.
This year our team competed in the 2018 Kaggle Data Science Bowl. The goal of the competition was to identify cells in microscopic images. There can be zero, one or many cells in any given image. Our solution was a Convolutional Neural Network with skip connections.
Although Pokemon Go! may have introduced the general public to augmented reality technology, it hasn’t come close to demonstrating the revolutionary impact AR will have on virtually everyone’s life. With numerous potential applications across a wide range of industries, AR is going to improve healthcare, make cities safer, and boost the efficiency of manufacturing processes (just to name a few examples).
Pipenv is a new tool for managing dependencies and virtualenv environments for Python projects.