國立虎尾科技大學 |

Prediction Problems Using Maximum Entropy Models.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	Prediction Problems Using Maximum Entropy Models./
作者:	Khribi, Lotfi.
面頁冊數:	1 online resource (100 pages)
附註:	Source: Dissertation Abstracts International, Volume: 79-07(E), Section: B.
Contained By:	Dissertation Abstracts International79-07B(E).
標題:	Statistics. -
電子資源:	click for full text (PQDT)
ISBN:	9780355649451

Prediction Problems Using Maximum Entropy Models.
Khribi, Lotfi.

Prediction Problems Using Maximum Entropy Models. - 1 online resource (100 pages)

Source: Dissertation Abstracts International, Volume: 79-07(E), Section: B.

Thesis (Ph.D.)--HEC Montreal (Canada), 2017.

Includes bibliographical references

In this thesis, we will study issues related to prediction problems. In particular, we consider the prediction of recurrent events. For this, we develop different prediction models when these events can be modeled using homogeneous or nonhomogeneous Poisson processes. Amongst these models, we are interested in those using random effects because they possess interesting features. We propose a predictive model using empirical Bayes techniques and the maximum entropy principle in order to model the random effects for the unknown parameters. We will show that for the prediction of recurrent events, our first model using as a prior the two moments maximum entropy distribution, which is equivalent to the truncated normal distribution, compared very favorably to the negative binomial model that uses as a prior the gamma distribution. We also present an extension of the approach developed in our first model: because of the two moment condition on our maximum entropy priors, we were restricted to considering only cases where the coefficient of variation was less than or equal to 1. We remove this restriction by the use of higher moment maximum entropy priors in the prediction of recurrent events using homogeneous and nonhomogeneous Poisson processes. We assess the performance of such models through extensive simulation studies and some real data sets.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2018

Mode of access: World Wide Web

ISBN: 9780355649451Subjects--Topical Terms:

556824
Statistics.
Index Terms--Genre/Form:

554714
Electronic books.

Prediction Problems Using Maximum Entropy Models.
LDR:02559ntm a2200325Ki 4500 001 920681
005 20181203094032.5
006 m o u
007 cr mn||||a|a||
008 190606s2017 xx obm 000 0 eng d
020 $a 9780355649451
035 $a (MiAaPQ)AAI10685735
035 $a (MiAaPQ)hec:10079
035 $a AAI10685735
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Khribi, Lotfi. $3 1195549
245 1 0 $a Prediction Problems Using Maximum Entropy Models.
264 0 $c 2017
300 $a 1 online resource (100 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertation Abstracts International, Volume: 79-07(E), Section: B.
500 $a Advisers: Marc Fredette; Brenda MacGibbon.
502 $a Thesis (Ph.D.)--HEC Montreal (Canada), 2017.
504 $a Includes bibliographical references
520 $a In this thesis, we will study issues related to prediction problems. In particular, we consider the prediction of recurrent events. For this, we develop different prediction models when these events can be modeled using homogeneous or nonhomogeneous Poisson processes. Amongst these models, we are interested in those using random effects because they possess interesting features. We propose a predictive model using empirical Bayes techniques and the maximum entropy principle in order to model the random effects for the unknown parameters. We will show that for the prediction of recurrent events, our first model using as a prior the two moments maximum entropy distribution, which is equivalent to the truncated normal distribution, compared very favorably to the negative binomial model that uses as a prior the gamma distribution. We also present an extension of the approach developed in our first model: because of the two moment condition on our maximum entropy priors, we were restricted to considering only cases where the coefficient of variation was less than or equal to 1. We remove this restriction by the use of higher moment maximum entropy priors in the prediction of recurrent events using homogeneous and nonhomogeneous Poisson processes. We assess the performance of such models through extensive simulation studies and some real data sets.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2018
538 $a Mode of access: World Wide Web
650 4 $a Statistics. $3 556824
655 7 $a Electronic books. $2 local $3 554714
690 $a 0463
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a HEC Montreal (Canada). $b Doctorat en administration. $3 1184797
773 0 $t Dissertation Abstracts International $g 79-07B(E).
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10685735 $z click for full text (PQDT)