Over the last couple of years, I had the opportunity to apply machine learning techniques to the financial sector. In this overview, I’ll show you how to use ARIMA modeling approaches to build your own financial models!
If you haven’t yet, set up your environment as described in this post!
A deep understanding of how ARIMA works is outside the scope of this post. For great theory resources, I recommend:
It is essential to have at least a basic understanding…
WaveNet is a powerful new predictive technique that uses multiple Deep Learning (DL) strategies from Computer Vision (CV) and Audio Signal Processing models and applies them to longitudinal (time-series) data. It was created by researchers at London-based artificial intelligence firm DeepMind, and currently powers Google Assistant voices.
This blog post accompanies a talk I recently gave at the Global AI Conference in Seattle (April 23–25, 2019). It also serves as a distillation of the Jupyter Notebook I used to give my lecture and lab, which can be found on my GitHub, along with supporting data and resources.
We’ll explore WaveNet…
I recently had the opportunity to spend six months in Seattle doing a teaching rotation with Metis, the data science school I work at. My first memories as a child were of exploring downtown and Lake Washington with my family; after many years living on the East Coast, I jumped at the chance to spend more time in the Emerald City. Now that I’ve had some time to reflect on my experience in Seattle, I’m coming to the realization that the city is one of the best places aspiring data scientists can call home.
I was excited when I first heard that Turi Create was acquired by Apple and then later open-sourced to the greater machine learning community! Earlier this year, I wrote about how Turi Create is Disrupting the Machine Learning Landscape. Then came WWDC18 and a host of improvements to Turi Create, including a beta version 5.0.
On my current research sabbatical with Metis, I’ve had the amazing opportunity to — long overdue — apply the data science and machine learning techniques that I teach daily to the financial sector. In this series of posts, I’ll show you how to get started building your own financial models!
First, I want to thank the team at Quandl for their amazing, easy-to-use platform, and the Quantopian community for great resources and inspiration!
This is by no means an “advanced” guide, and while here, I should mention:
The information provided here and accompanying material is for informational purposes only. It…
At Metis, one of the first machine learning models I teach is the Plain Jane Ordinary Least Squares (OLS) model that most everyone learns in high school. Excel has a way of removing the charm from OLS modeling; students often assume there’s a scatterplot, some magic math that draws a best fit line, then an r² in the corner that we’d like to get close to 1. Truth is, there’s so much more than meets the eye with OLS, and after about a week, students are crying for mercy (disclaimer: no students are actually harmed!) …
Below is a distilled collection of conversations, messages, and debates I’ve had with peers and students on how to optimize deep models. If you have tricks you’ve found impactful, please share them!!
Deep learning models like the Convolutional Neural Network (CNN) have a massive number of parameters; we can actually call these hyper-parameters because they are not optimized inherently in the model. You could gridsearch the optimal values for these hyper-parameters, but you’ll need a lot of hardware and time. So, does a true data scientist settle for guessing these essential parameters?
One of the best ways to improve your…
At the end of 2017, Apple made a splash in the machine learning domain when it released Turi Create, an open source package that makes it easier for developers to put the power of machine learning into their apps. The company’s announcement claimed Turi Create would allow users to “focus on tasks instead of algorithms.” Let’s face it: that’s music to the ears of business leaders with limited time and resources!
Now that the data scientists and software developers of the world have had some time to absorb the news and start working with Turi Create, let’s break down how…
Deep learning neural networks — with more than one hidden layer — require a dumptruck load of data to become effective predictive engines. These models require powerful computing nodes and big RAM stores, and loads of data need ample storage and caching capabilities.
With the price and ease of use that cloud solutions like Amazon Web Services (AWS) bestow, it doesn’t make sense to drop the heavy burden of training on your precious laptop!
Plus, free up your local machine to browse Twitter or watch data science tutorials while your model converges remotely!
This guide will get you:
A few months ago — Thursday, January 18 — I presented at my first Seattle meetup. This was three days after I moved my family cross-country from Chicago, so no time was wasted!
Bottom line: Turi Create is a Pythonic Machine Learning library that’s amazingly powerful and easy to use, and you should explore its capabilities!