Twitter Weekly Summary for 2009-10-17

  • back from long weekend… updating my forex positions #
  • just wasted a few minutes on Financial Soccer – http://bit.ly/6JOKC #
  • Dr. Moore's useful statistical data mining tutorials… http://bit.ly/ovsZR #
  • not looking at market much today… GBP/USD looks like a potential short though #
  • finished Good to Great this weekend #
  • looked at self publishing book and ebook #
  • EUR/USD double top in the making? #
  • shorted NZD/USD @ .73814 #
  • considered going long USD/CAD, but the new low is too intimidating, so I went for NZD/USD with a lower higher #
  • moved EUR/JPY stop up to 132.00 #
  • brushing up on probability analytics for data mining…http://www.autonlab.org/tutorials/prob18.pdf #
  • reading up on multivalued random variables… it's been a few years since I've studied these stuff #
  • Bayesian inference…http://bit.ly/kM5xg #
  • 3 major models for data mining: classifier, density estimator, regressor #
  • I just need to understand the concept, no need to know the details like proving theorems, understanding equations, etc. #
  • idea: use a density estimator to detect data anomaly #
  • on slide 49 of http://www.autonlab.org/tutorials/prob18.pdf about probability analytics, will continue tomorrow #
  • my NZD/USD short is stopped out again! #
  • couldn't login my forex trading platform for some reason #
  • got in my forex trading platform, it was my browser's problem #
  • will wait for NZD/USD to break short-term support around 0.7300 before shorting again. #
  • EUR/JPY exhausting… moved up stop again to 132.32 #
  • if my EUR/JPY long gets stopped out, I'll look for USD strength #
  • not neglect that fact that EUR/USD could spike above 1.49 for a last run…not playing for USD strength just yet #
  • stop entry long EUR/USD 1.4921, just above today's high. Tight stop. 1/2 position. Try for a breakout above 1.49. #
  • DOW 10,000 merely 20 pts away #
  • no EUR/USD spike yet #
  • looks as though everything is at standstill waiting for DJIA 10,000 #
  • DJIA countdown? 10..9… 8… #
  • so much fuzz about DJIA hitting 10,000 from the media #
  • and it lasted only seconds #
  • let's see if we can CLOSE above 10,000 this week #
  • in EUR/USD at 1.4921, going for the obvious #
  • that was for a EUR/USD long #
  • back to the probability analytics slides #
  • keeping away from the market for now by not entering new positions as we're in a new territory #
  • finished slides on probability analytics, but merely skimped through last part about Bayesian Classifiers http://www.autonlab.org/tutorials/ #
  • EUR/USD and EUR/JPY edging up #
  • launching #RapidMiner 5 to fiddle with it after a few days in the shelf #
  • my computer just made some weird noise… backing up some stuff now #
  • imported SPY data since 1995 into #RapidMiner… just playing around with it for now #
  • need to reassess my EUR/JPY position for the target #
  • moved take profit stop on my EUR/JPY long to 136.00, since we're approaching 134 already and my prev. 135 stop is too close. #
  • trying a Support Vector Clustering model on my SPY data #
  • I need to figure out how to do Time Series analysis using #RapidMiner #
  • 5 min. in and #RapidMiner is still churning my model. My backup/compression is really taking a lot of CPU time away #
  • forgot to add a model writer to save the model. so need to restart now #
  • tried to add a X-validator to the model. but realized that it wouldn't apply on clustering since there's no label to test #
  • I'm running some data and CPU time only indicates 1-core used. #RapidMiner doesn't appear to be multi-core capable? #
  • according to the company site, the free version of #RapidMiner is indeed single-core only #
  • still long EUR/JPY from 130.00, raised stop to 133.50 and reviewing my entry journal to see where I should aim as target #
  • moved EUR/JPY stop to 134, my original target. then moved take profit stop to 137 from 136 to let it ride. #
  • EUR/JPY above 135! Let's see how long it can hold. #
  • if #Goldman (GS) can't fill the gap, this could mean trouble for the stock market #
  • watching GS and AAPL as indicator for the market today #
  • downloading #RapidMiner for my work pc so that I can run the data analysis overnight when needed #
  • just realized I have a single-core Pentium 4 @ 3.2Ghz at work… it could be faster than my PC at home for running #RapidMiner CE analysis! #
  • Yahoo Pipes look like a useful tool to fetch data from different sources for data mining http://pipes.yahoo.com/pipes/ #
  • I can use Yahoo Pipes to fetch and aggregate up-to-date quotes, then import it direct from #RapidMiner! Saves me from daily manual updates! #
  • Here's a term which I've just learned — data scraping. It's what I've been griefing over these days. http://bit.ly/1j8rdY #
  • here's an idea — use Pipes to fetch individual quotes from a stock index, then feed into #RapidMiner to reverse-engineer the algorithm #
  • USD/JPY's failure to break 91.00 is a concern for my EUR/JPY long #
  • although that was just a first try and it haven't fallen back down yet #
  • keeping my EUR/JPY stop at 134.50, but placing an order to add 100% more at 134.80 with a tight stop for that also at 134.50 #
  • suddenly a lot of price movement in the forex market #
  • … which lasted a few seconds #
  • long USD/CAD for 1/4 position just below 1.0300, stop at the recent low #
  • will write about my USD/CAD long entry tomorrow morning #
  • well, my USD/CAD long worked out well with +100 pips now, I've moved my stop up to breakeven and will add if we make a higher low #
  • #FF @BladeTrading trades #forex night and day #
  • moved EUR/JPY stop up to 135.00 because it's friday #
  • #FF @Ancient_Warrior trades #forex and #stocks #
  • back to work #
  • stopped out of EUR/JPY at 135.00 as expected for a +500 pips… better to sleep safe over the weekend than wait for 138. Can try later. #
  • will analyze my EUR/JPY trade and USD/CAD long entry this weekend. need to catch up on day job now #
  • another hour wasted tweaking the website #
  • updating my trade performance stats #
  • finally done integrating my trade stats on my blog, took a few good hours http://www.quantisan.com/trade-log/ #
  • #beatcancer when is the deadline for this? #
  • #beatcancer seems that this is just over http://beatcancereverywhere.com/about.html #
  • I'm so out of shape in playing video games after so long… can't even pass level 1 of Resident Evil 5 on normal…now where's that cheat? #
  • looking to create a gantt chart to log my quant research to keep myself motivated and to keep track #
  • Just added myself to the http://wefollow.com twitter directory under: #ottawa_o #forex #quant #investing #stocks #trad #
  • Just added myself to the http://wefollow.com twitter directory under: #ottawa #investing #stocks #trading #forex #quant #
  • backed up website and working logging timesheet #
  • built my first pipe to scrape yahoo finance historical stock quote in .csv http://pipes.yahoo.com/quantisan/stockhistoricalcsv #

Related posts:

  1. Twitter Weekly Summary for 2009-10-04
  2. Twitter Weekly Summary for 2009-10-10
  3. Twitter Weekly Summary for 2009-11-28
  4. Twitter Weekly Summary for 2009-10-25
  5. Twitter Weekly Summary for 2009-11-01

Leave a Reply

Your email address will not be published. Required fields are marked *

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>