There are countless open-source optimization packages that can help you minimize an arbitrary multivariate function, even if you don’t know the derivatives of that function. They are a lot better than grid or random search, but where should you look first?
I provide a simple answer in the form of a colab notebook that you can easily modify for your purposes. This post provides a walk-through and some background. But even before opening that notebook, you might want to browse the Elo ratings for global optimizers which inform the same. Here are some strategies that perform well on my problems.
The popular open-source PyCaret package provides automated machine learning capability, allowing the user to search hundreds of regression models. The TimeMachines package provides a variety of incremental (online) time-series algorithms. In this short post, we cover the nuts and bolts of using these two libraries together. We shall
All code is provided in https://github.com/microprediction/timeseries-notebooks/blob/main/pycaret_microprediction_timemachines.ipynb. Let’s get a few preliminaries out of the way.
!pip install pycaret[full]
!pip install --upgrade statsmodels
!pip install microprediction
!pip install timemachines
from datetime import datetime, timedelta
from microprediction import MicroReader
As a kid, one of my introductions to applied mathematics was the racetrack. Specifically, it was Randwick Racecourse in Sydney. I won’t bore you with my various adventures, except to say that for the longest time I have felt a strong pull towards a mathematical problem that arises there — one that seemed not to be well covered in the order statistics literature. I’d go so far as to say I had an overwhelming aesthetic need to construct a self-consistent pricing model for all racetrack wagers.
I’m pleased to say that some work down these lines was recently published in…
Graham Giller has been a data scientist for a lot longer than that phrase has been in existence. In a new book, he reflects on some of his undertakings— which involved working with one of Wall Street’s best-known quantitative trading operations.
The work falls into a genre that is probably underserved. It is rare that “non-famous” people, as Graham refers to himself in the book, take the time to chronicle their efforts. Yet it seems highly unlikely that aspiring data scientists could not learn something from the experience of others.
I asked Graham some questions about the work, which begins…
I suggest you try this on every financial series at www.microprediction.org before drawing any conclusions. Here you are using a very small sample.
Quite a few open source optimization packages were reviewed in a previous blog article Comparing Python Global Optimizers. Here I introduce a small Python package I’ve been writing, called embarrassingly, that might work in tandem with any one of them. The library helps you modify your objective function before you send it to an optimizer, and this article explains why you might want to do that sometimes.
Here’s an example. I want to land my helicopter on the red plateau, even though there are higher points to choose further from the origin (we’re minimizing, but the plot is upside down…
Facebook’s Prophet package aims to provide a simple, automated approach to the prediction of a large number of different time series. The package employs an easily interpreted, three-component additive model whose Bayesian posterior is sampled using STAN. In contrast to some other approaches, the user of Prophet might hope for good performance without tweaking a lot of parameters. Instead, hyper-parameters control how likely those parameters are a priori, and the Bayesian sampling tries to sort things out when data arrives.
Judged by popularity, this is surely a good idea. Facebook’s prophet package has been downloaded 13,698,928 times according to pepy…
Can NFL teams win more games by instructing the ball carrier to stop short of the first down? Believe it.
This year marks the 100th anniversary of the National Football League. The game has undergone many changes, but statistical analysis of strategy has been something of a late starter. For example, a good chunk of that history expired before a serious discussion of punting strategy took place (to this day, some armchair statisticians are driven to distraction by suboptimal fourth-down decision making).
In this post, I go after another sacred cow — the first down. I’m going to…
Chief Data Scientist, Intech Investments