This work proposes a sequential methodology for selecting variables in classification problems in which the number of predictors is much larger than the sample size. The methodology includes a Monte Carlo permutation procedure that conditionally tests the null hypothesis of no association among the outcomes and the available predictors. In order to improve computing aspects, we propose a new parametric distribution, the Truncated and Zero Inflated Gumbel Distribution. The final application is to find compact classification models with improved performance for genomic data. Results using real data sets show that the proposed methodology selects compact models with optimized classification performances.
MacCullaghPNelderJA. Generalized linear models. Monographs on statistics and applied probability, London: Chapman and Hall, 1989.
2.
TibshiraniR. Regression shrinkage and selection via the lasso. J R Stat Soc Ser B1996; 58: 267–288.
3.
FriedmanJHHastieTTibshiraniR. Regularization paths for generalized linear models via coordinate descent. J Stat Softw2012; 33(1): 1–22.
4.
ZareHHaffariGGuptaA, et al.Scoring relevancy of features based on combinatorial analysis of lasso with application to lymphoma diagnosis. BMC Genom2013; 14(Suppl 1): S14–S14.
5.
ZouHHastieT. Regularization and variable selection via the elastic net. J R Stat Soc: Ser B2005; 67(2): 301–320.
MontgomeryDCPeckEAViningGG. Introduction to linear regression analysis, 5th ed. New Jersey: Wiley, 2012.
8.
DobsonAJ. An introduction to generalized linear models, London: Chapman and Hall, 2008.
9.
AbramsAMKleinmanKKulldorffM. Gumbel based p-value approximations for spatial scan statistics. Int J Health Geogr2010; 9(61): 1–12.
10.
KulldorffM. A spatial scan statistic. Commun Stat Theory Methods1997; 26: 1481–1496.
11.
CostaMAAssunçãoRMKulldorffM. Constrained spanning tree algorithms for irregularly-shaped spatial clustering. Comput Stat Data Anal2012; 56(6): 1771–1783.
12.
DwassM. Modified randomization tests for nonparametric hypotheses. Ann Math Stat1957; 28: 181–187.
13.
DudoitSYangYCallowM, et al.Statistical methods for identifying differentially expressed genes in replicated cDNA microarray experiments. Stat Sin2002; 12(1): 111–140.
14.
WaldAWolfowitzJ. Statistical tests based on permutations of the observations. Ann Math Stat1944; 15(4): 358–372.
15.
YpmaTJ. Historical development of the Newton–Raphson method. Soc Ind Appl Math1995; 37: 531–551.
16.
FriedmanJHastieTTibshiraniR. Regularization paths for generalized linear models via coordinate descent. J Stat Softw2010; 33(1): 1–1.
17.
MazumderRFriedmanJHHastieT. Sparsenet: coordinate descent with nonconvex penalties. J Am Stat Assoc2011; 106(495): 1125–1138.
18.
ZhangC-H. Nearly unbiased variable selection under minimax concave penalty. Ann Stat2010; 38(2): 894–942.
19.
Bach FR. Bolasso: model consistent lasso estimation through the bootstrap. In: Proceedings of the 25th international conference on Machine learning. New York: ACM, 2008, pp.33–40.
20.
HaykinS. Neural networks and learning machines, 3rd ed. New Jersey: Prentice Hall, 2008.
21.
KlebanovLYakovlevA. How high is the level of technical noise in microarray data. Biol Direct2007; 2(9): 1–9.
22.
HessKAndersonKSymmansW, et al.Pharmacogenomic predictor of sensitivity to preoperative chemotherapy with paclitaxel and fluorouracil, doxorubicin, and cyclophosphamide in breast cancer. J Clin Oncol2006; 24(26): 4236–4244.
23.
NatowiczRBragaAIncittiR, et al.A new method of DNA probes selection and its use with multi-objective neural network for predicting the outcome of breast cancer preoperative chemotherapy. ESANN2008. https://www.elen.ucl.ac.be/Proceedings/esann/esannpdf/es2008-63.pdf.
24.
NatowiczRIncittiRHortaEG, et al.Prediction of the outcome of preoperative chemotherapy in breast cancer by dna probes that convey information on both complete and non complete responses. BMC Bioinform2008; 9(149): 71–76.
25.
NatowiczRIncittiRRouzierR, et al.Downsizing multigenic predictors of the response to preoperative chemotherapy in breast cancer. Lecture notes in computer science, New York: Springer, 2008, pp. 157–164.
26.
Bolstad, Benjamin M., et al. A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics 2003; 19(2): 185–193.
27.
van ’t VeerLJDaiHvan de VijverMJ, et al.Gene expression profiling predicts clinical outcome of breast cancer. Nature2002; 415(6871): 530–536.
28.
SakamotoYIshiguroMKitagawaG. Akaike information criterion statistics, New York: Springer, 1999.
29.
CostaMAAssunçãoRM. A fair comparison between the spatial scan and the besag–newell disease clustering tests. Environ Ecol Stat2005; 12(3): 301–319.
30.
TangoTTakahashiK. A flexibly shaped spatial scan statistic for detecting clusters. Int J Health Geogr2005; 4(1): 11–11.
31.
WaldA. Sequential analysis, New York: Courier Corporation, 1973.
32.
SiegmundD. Sequential analysis: tests and confidence intervals, New York: Springer, 1985.
33.
EfronBHastieTJohnstoneI, et al.Least angle regression. Ann Stat2004; 32(2): 407–499.
34.
ChengPChengYLiY, et al.Comparison of the gene expression profiles between smokers with and without lung cancer using rna-seq. Asian Pac J Cancer Prev2012; 13: 3605–3609.
35.
PanditTSKennetteWMacKenzieL, et al.Lymphatic metastasis of breast cancer cells is associated with differential gene expression profiles that predict cancer stem cell-like properties and the ability to survive, establish and grow in a foreign environment. Int J Oncol2009; 35(2): 297–297.
36.
WangXSouthardRCAllredCD, et al.Maz drives tumor-specific expression of ppar gamma 1 in breast cancer cells. Breast Cancer Res Treat2008; 111(1): 103–111.
37.
ClarkJEdwardsSJohnM, et al.Identification of amplified and expressed genes in breast cancer by comparative hybridization onto microarrays of randomly selected cdna clones. Genes Chromosomes Cancer2002; 34(1): 104–114.
38.
RamaswamySRossKNLanderES, et al.A molecular signature of metastasis in primary solid tumors. Nature Genet2002; 33(1): 49–54.
39.
BiècheILereboursFTozluS, et al.Molecular profiling of inflammatory breast cancer identification of a poor-prognosis gene expression signature. Clin Cancer Res2004; 10(20): 6789–6795.
40.
RamaswamyBMajumderSRoyS, et al.Estrogen-mediated suppression of the gene encoding protein tyrosine phosphatase ptpro in human breast cancer: mechanism and role in tamoxifen sensitivity. Mol Endocrinol2009; 23(2): 176–187.
41.
HenryJLCogginDLKingCR. High-level expression of the ribosomal protein l19 in human breast tumors that overexpress erbb-2. Cancer Res1993; 53(6): 1403–1408.
42.
XieDJauchAMillerCW, et al.Discovery of over-expressed genes and genetic alterations in breast cancer cells using a combination of suppression subtractive hybridization, multiplex fish and comparative genomic hybridization. Int J Oncol2002; 21(3): 499–507.
43.
SloanDDNicholsonBUrquidiV, et al.Detection of differentially expressed genes in an isogenic breast metastasis model using rna arbitrarily primed-polymerase chain reaction coupled with array hybridization (rap-array). Am J Pathol2004; 164(1): 315–323.
44.
McCaigCPerksCMHollyJM. Intrinsic actions of igfbp-3 and igfbp-5 on hs578t breast cancer epithelial cells: inhibition or accentuation of attachment and survival is dependent upon the presence of fibronectin. J Cell Sci2002; 115(22): 4293–4303.
