Disqus - Latest Comments for ericbrown

Re: Book: Tulsa 1921

Eric D. Brown — Sun, 12 Jul 2020 21:42:06 -0000

I grew up in Oklahoma in the 70's and 80's. Graduated High School in '91 and went to college in Oklahoma. I can't recall one minute spent in any history classes on the Tulsa Race Riot. I'm a history buff and stumbled upon it on my own and was blown away by the stories and reality of what happened. I lived in Tulsa for a few years and regularly went down to the Greenwood area and walked the streets where Black Wall Street once stood. There are a plaques along the street describing what had been in that spot before the destruction.

Its definitely a 'white-washed' piece of history that many have never heard.

Re: Algolia

Eric D. Brown — Mon, 09 Dec 2019 11:19:36 -0000

I've been using Algolia for about 2 years now and absolutely love it. If you're a developer, you can customize results and search criteria and the service is lightning fast. Algolia is one of my rare 'very highly recommenced' services.

Re: Pro Wildlife Photographer Takes Us to Wild Alaska with the Sony a9 – Real World Review

Eric D. Brown — Sat, 20 Oct 2018 11:23:06 -0000

He published this video on August 18 2018, before the Z release

Re: You Married Facebook

Eric D. Brown — Tue, 08 May 2018 09:23:53 -0000

Same goes for Twitter, Instagram, Youtube, Google, Snapchat, Apple, Amazon...etc etc. The list goes on.

Everything you do, everywhere you go and everyone you know is easily accessible by any/all of these sites.

Deleting Facebook isn't going to make that much of a difference in how easily you are 'fucked with'.

Re: Technology can be a band-aid rather than a solution

Eric D. Brown — Fri, 28 Jul 2017 08:51:55 -0000

I agree. Thanks for stopping by.

Re: New Feature – Backtesting

Eric D. Brown — Wed, 26 Jul 2017 10:35:23 -0000

Hi Ali - We are planning on doing that. Should be on the page when I push a new version over the weekend.

Re: New Feature – Backtesting

Eric D. Brown — Wed, 26 Jul 2017 08:20:34 -0000

Sorry...missed that part of your question.

The backtester only looks at days where trading data is available, so the observation days is based on trading days.

Good suggestion on the enhancement. I've been thinking about doing something that matches up with the daily reports...I've added this to the enhancement list.

Re: New Feature – Backtesting

Eric D. Brown — Wed, 26 Jul 2017 06:38:03 -0000

The lookback period is the amount of history you want the backtester to look at. For example, if you select 2 years, it will load 2 years of history to use in backtesting.

The observation period is how much time 'in the future' you want to look at the returns. For example, if you select 10 days, the backtester will provide the returns 10 days beyond the 'signal date'.

Re: I am now a doctoral student

Eric D. Brown — Tue, 11 Jul 2017 17:34:23 -0000

I chose DSS because that's what I was (and still am) interested in.

I can't really recommend any of the specialties over another unless I know what your interests are. If you are more interested in Cyber Security, DSU has a top-notch program. They have some partnerships with the NSA and other large organizations.

Re: You (probably) don’t need Machine Learning

Eric D. Brown — Thu, 18 May 2017 09:54:44 -0000

Thanks Brian. Well said. Hopes aren't high for me either, but all we can do is keep trying educate and inform.

Re: You (probably) don’t need Machine Learning

Eric D. Brown — Thu, 13 Apr 2017 23:07:58 -0000

Well said Heidi. Love the analogy as well.

Re: You (probably) don’t need Machine Learning

Eric D. Brown — Tue, 11 Apr 2017 12:38:40 -0000

Thanks Eric.

This is a good description of the requirements needed before you get to ML. Lots of data cleaning, management, etc needed.

Re: You (probably) don’t need Machine Learning

Eric D. Brown — Sun, 15 Jan 2017 09:52:16 -0000

Exactly Thomas.

Now...if they have their data cleaned, organized and ready to go AND they've tried everything else (or their use case is one of the ones that really fits ML), then maybe - just maybe - they can try out ML.

Re: You (probably) don’t need Machine Learning

Eric D. Brown — Fri, 06 Jan 2017 08:34:07 -0000

correct.

Re: You (probably) don’t need Machine Learning

Eric D. Brown — Thu, 05 Jan 2017 07:41:28 -0000

Good old fashioned regression modeling is a statistical method that has been around much longer than machine learning. Regression is now used within machine learning methods and processes but is wholly separate from machine learning as a modeling technique (in its purest form).

Thanks for the question.

Re: You (probably) don’t need Machine Learning

Eric D. Brown — Wed, 04 Jan 2017 15:30:18 -0000

Well said Bill.

Re: Stock Indicators at an Extreme – Dec 20 2016

Eric D. Brown — Thu, 22 Dec 2016 15:34:12 -0000

Thanks Enrico. I've made some notes on the requests and will take a look at them.

Re: Following the Smart Money (Part 3)

Eric D. Brown — Thu, 17 Nov 2016 08:16:52 -0000

I use a combination of Tradestation and python. I do most of the work in python (I use backtrader - https://github.com/mementum... and I then use Tradestation to 'confirm' the python tests. The statistics and charts you see are outputs from pyfolio (https://github.com/quantopi....

Re: Following the Smart Money

Eric D. Brown — Sat, 29 Oct 2016 20:54:26 -0000

No worries on the question.

I had been doing all strategies with 1000 shares to make things simpler in backtesting but adding the variable shares to the strategies turned out to be fairly easy to do. Going forward, I'll be using the variable shares approach.

Re: Hidden Markov Models for Regime Detection using R

Eric D. Brown — Mon, 03 Oct 2016 16:43:36 -0000

I actually agree. I see this type of work as something that is interesting to have to get a feel for what regime we might be in 'now' (based on retrained data, etc) but not for any predictive capabilities.

Re: Hidden Markov Models for Regime Detection using R

Eric D. Brown — Mon, 03 Oct 2016 16:32:14 -0000

Ilya -

I know you write about R a lot but don't recall any regime change work you've done. Do you have any recommendations for a better approach for live/online predictions?

Re: Stock Indicators at an Extreme – Sept 8 2016

Eric D. Brown — Thu, 08 Sep 2016 11:48:40 -0000

Hi Brian -

The model_percent_spread is similar (it is the spread between model_percent_bullish and model_percent_bearish). In the tables above, there might be some indicators that are not included in the model calculations but are provided above.

Re: Social Sentiment for select ETF’s

Eric D. Brown — Tue, 09 Aug 2016 11:14:33 -0000

Hi Bill -

Correct. The lower the value, the more bullish the sentiment and a change for correction at the extremes on the low end.

Re: Strategy Development with the SPY Optimism Index (Optix)

Eric D. Brown — Wed, 13 Jul 2016 08:24:01 -0000

Glad you like it.

I am actually going to be looking at the Short and Intermediate models in coming articles.

Re: Strategy Development with the SPY Optimism Index (Optix)

Eric D. Brown — Tue, 12 Jul 2016 21:32:24 -0000

Thanks Ken. Glad you like it.