A complete, step by step guide to building a production-grade machine learning app with Django, PostgreSQL, React, Redux and Docker
Archives for 2020
Forecasting time series is important in many contexts and highly relevant to machine learning practitioners. Take, for example, demand forecasting from which many use cases derive. Almost every manufacturer would benefit from better understanding demand for their products in order to optimise produced quantities. Underproduce and you will lose revenues, overproduce and you will be forced to sell excess produce at a discount. Very related is pricing, which is essentially a demand forecast with a specific focus on price elasticity. Pricing is relevant to virtually all companies.
This tutorial demonstrates porting an existing machine learning model to a virtual machine on the Microsoft Azure cloud platform. We will train a small movie recommendation model using a single GPU to give personalised recommendations. The total cost of performing this training should be no more than $5 using any of the single GPU instances currently available on Azure.
Training is without a doubt the most important part of developing a machine learning application. It’s when you start realizing whether or not your model is worth it, how your hyperparameters should look like and what do you need to change in your architecture. In general, most machine learning engineers spend quite some time on training, experimenting with different models, tuning their architecture and discovering the best metrics and losses for their problem.
We’ll demonstrate the usage of concurrent HTTP requests by fetching prices for stock tickers. The only third party package we’ll use is httpx. Httpx is very similar to the popular requests package, but httpx supports asyncio.
This book covers the building blocks of the most common methods in machine learning. This set of methods is like a toolbox for machine learning engineers. Those entering the field of machine learning should feel comfortable with this toolbox so they have the right tool for a variety of tasks. Each chapter in this book corresponds to a single machine learning method or group of methods. In other words, each chapter focuses on a single tool within the ML toolbox.
In order to implement an algorithmic trading strategy though, you have to first narrow down a list of stocks that you want to analyze. This walk-through provides an automated process (using python and logistic regression) for determining the best stocks to algo-trade.
I will dive deeper into the logic and code below, but here is a high-level overview of the process:
- Import the historical data of every stock using yahoo finance.
- Pull in over 32 technical indicators for each stock using the technical analysis library.
- Perform a logistic regression on each stock using 5, 30, and 60 day observation time periods.
- Interpret the results.
Array programming provides a powerful, compact and expressive syntax for accessing, manipulating and operating on data in vectors, matrices and higher-dimensional arrays. NumPy is the primary array programming library for the Python language. It has an essential role in research analysis pipelines in fields as diverse as physics, chemistry, astronomy, geoscience, biology, psychology, materials science, engineering, finance and economics.
In this tutorial, you will be learning how to build powerful time-series forecasting model of your own using various kinds of deep learning algorithms such as Dense Neural Networks (DNN), Convolutional Neural Network (CNN) and Recurrent Neural Networks (RNN). Also, this course is an elaboration of the time-series forecasting tutorial by TensorFlow.
LendingClub is the world’s largest peer-to-peer lending platform. Until recently (through the end of 2018), LendingClub published a public dataset of all loans issued since the company’s launch in 2007.
This post reviews NumPy main components and functionality, with attention to the needs of Data Science and Machine Learning practitioners, and people who aspire to become a data professional.
Lessons learned building a profitable algorithmic trading system using Reinforcement Learning techniques.
The plotting functionality in the popular Python data analysis library Pandas has always been one of my go-to methods for super quick charts. However, the available visualisations have always been fairly basic and not particularly pretty.
It’s easy to get carried away with the wealth of data and free open-source tools available for data science. After spending a little bit of time with the quandl financial library and the prophet modeling library, I decided to try some simple stock data exploration. Several days and 1000 lines of Python later, I ended up with a complete stock analysis and prediction tool. Although I am not confident (or foolish) enough to use it to invest in individual stocks, I learned a ton of Python in the process and in the spirit of open-source, want to share my results and code so others can benefit.
MlFinlab is a python package which helps portfolio managers and traders who want to leverage the power of machine learning by providing reproducible, interpretable, and easy to use tools. Adding MlFinLab to your companies pipeline is like adding a department of PhD researchers to your team.
The price of energy changes hourly, which opens up the possibility of temporal arbitrage: buying energy at a low price, storing it, and selling it later at a higher price. To successfully execute any temporal arbitrage strategy, some amount of confidence in future prices is required, to be able to expect to make a profit. In the case of energy arbitrage, the constraints of the energy storage system must also be considered. For example, batteries have limited capacity, limited rate of charging, and are not 100% efficient in that not all of the energy used to charge a battery will be available later for discharge.
This series explains concepts that are fundamental to deep learning and artificial neural networks for beginners. In addition to covering these concepts, we also show how to implement some of the concepts in code using Keras, a neural network API written in Python. We will learn about layers in an artificial neural network, activation functions, backpropagation, convolutional neural networks (CNNs), data augmentation, transfer learning and much more!
Exchange rates API is a simple and lightweight free service for current and historical foreign exchange rates.
In this post we will discuss how to do a time series modelling using ARMA and ARIMA models. Here AR stands for Auto-Regressive and MA stands for Moving Average
Since its emergence in Asia late 2019, the coronavirus COVID-19 pandemic has been devastating. The virus spread to most countries causing severe respiratory infections and many human casualties. The virus also put half of the world population in lockdown which resulted in a slowdown of the world economy and a fall in stock prices.
The goal of this tutorial is to introduce the steps for collecting and analyzing stock data in the context of the coronavirus pandemic. To do this, we will use Python, Google Sheets and Google Finance.
In finance, computation efficiency can be directly converted to trading profits sometimes. Quants are facing the challenges of trading off research efficiency with computation efficiency. Using Python can produce succinct research codes, which improves research efficiency. However, vanilla Python code is known to be slow and not suitable for production. In this post, I explore how to use Python GPU libraries to achieve the state-of-the-art performance in the domain of exotic option pricing.
If you have a relative working in the banking industry, ask the person what annoys him/her most about the job. You will surely receive an answer that is related to the task of data entry i.e. the practice of manually entering serial numbers and names from financial documents into the bank’s database.
Lots of quantitative risk metrics for analyzing your backtest and trading performance. Created by Quantopian for their popular Zipline backtesting framework, this library works totally independently.
The experience while accessing the AI platform and running machine learning (ML) training code on the platform must be smooth and easy for the researchers. Migrating any ML code from a local environment to the platform should not require any refactoring of the code at all. Infrastructure configuration overhead should be minimal. Our mission while developing PyKrylov was to abstract the ML logic from the infrastructure and Krylov core components as much as possible in order to achieve the best experience for the platform users.
One of the milestones of the investment management application was to implement an end to end solution that starts by fetching company stock prices and builds a set of efficient and optimum portfolios using optimisation routines.
Predictive model to correctly forecast future trend is crucial for investment management and algorithmic trading. The use of technical indicators for financial forecasting is quite common among the traders. Input window length is a time frame parameter required to be set when calculating many technical indicators.
Our problem here is to define whether or not a certain news article is fake news. The dataset is comprised of 3997 news articles each includes a title, text, and the target label as a REAL/FAKE binary label. Part of the course was also testing the model on a test dataset but I never received target for this dataset. The accuracy score of cross validation testing within the training dataset was 94%.