Tom Gastaldi - Selected original works

Distribution Theory: multinomial tails (G-N conjecture)

  On the minimization of multinomial tails and the Gupta Nagel conjecture (Multinomial distribution, Selection Procedures) [preliminary draft]

Prof. Tommaso Gastaldi: Published version (JMVA - Journal of Multivariate Analysis) : 
http://www.sciencedirect.com/science/article/pii/S0047259X04002180  [published article]

94-9C. T. Gastaldi, S. S. Gupta. Minimax Type Procedures for Nonparametric Selection of the "Best" Population with Partially Classified Data. [pdf]

From: http://www.stat.purdue.edu/research/technical_reports/1994-tr.html

http://en.scientificcommons.org/tommaso_gastaldi

Non-parametric sequential test G-SKS by Tom Gastaldi (IEEE Transaction on Reliability)

http://ieeexplore.ieee.org/Xplore/login.jsp?url=http%3A%2F%2Fieeexplore.ieee.org%2Fiel1%2F24%2F9132%2F00406587.pdf%3Farnumber%3D406587&authDecision=-203

(The name of the new test "G-SKS" was proposed by the Editor of the Journal)

Polynomial Regression for Fuzzy Data by Tom Gastaldi (Fuzzy Sets and Systems)

http://dl.acm.org/citation.cfm?id=766285
http://academic.research.microsoft.com/Publication/14496751/a-least-squares-approach-to-fuzzy-linear-regression-analysis
http://academic.research.microsoft.com/Publication/732513/an-orderwise-polynomial-regression-procedure-for-fuzzy-data

Special Kolmogorov-Smirnov test for arbitrarily censored data

http://www.tandfonline.com/doi/abs/10.1080/03610929308831004
http://academic.research.microsoft.com/Search?query=tommaso%20gastaldi%20censored



Multi-component systems and masked data
http://academic.research.microsoft.com/Publication/16185668/improved-maximum-likelihood-estimation-for-component-reliabilities-with




Analysis of "variance" for qualitative data ("mutability analysis")

A decomposition of an appropriate qualitative variation index (essentially a measure of entropy) is pointed out
(this is similar to the well known decomposition of variance, but for nominal variables):
Total Variability = Variability within groups + Variation between groups
and, most importantly, a simple and suitable distribution of the test statistics is derived.
This allows to treat quantitative data with a conceptual approach analogous to quantitative data, and the test is very simple and effective to carry out.

http://serials.unibo.it/cgi-ser/start/it/spogli/ds-s.tcl?authors=%22T.Gastaldi%22&language=ITALIANO
http://www.biblio.liuc.it/scripts/essper/ricerca.asp?tipo=autori&codice=2005394


Some proposals for algorithmic trading metrics

http://www.datatime.eu/public/gbot/MetricsForAlgorithmicTrading.htm

Author of:

 DT Universal Rolap System [Universal multidimensional ROLAP + quaternion engine]

 The G-BOT Project   [Algorithmic Trading Methodologies - Fund automated management]   (active)

Linkedin groups:

"Algorithmic Trading for Real Traders and Funds" : http://www.linkedin.com/groups?gid=4394344&trk=hb_side_g

"Interactive Brokers Traders and Fund Managers" http://www.linkedin.com/groups/Interactive-Brokers-Traders-Fund-Managers-4932221

(If interested in full CV, please write: tommaso.gastaldi@gmail.com)

Contact: