ADS Capstone Chronicles Revised
18
Ac ol tmh op uo gnhe n9t s0, (ma )a iyt isst il lel s ss ot hu an nd al itkhei r qd uoi ft et hae 2f e7w4 SbDecOaHusceolumns we had prior to PCA and (b) we have 200,583 training of ebasteurrveast ii no nosu, r fwe ae t uar rees e at bwl eh i l et os t i hl l abveei n gmwa ne lyl ambaoi vnet a i ntihneg a twl ee al l s- et s1t 0a bol ibssheerdv a t igounisd ef ol irn ee v e royf 11 9f 9e a6t)u. rTeh ui ns , awme oc do enldi nugc tdPaCt aAs oe tn (oPuerd 2u 7z z4i SeDt Oa Hl . , cS oi ml ui ml a nr lsy ttoo bt hr iensgc iat l idnogwann dt oi m9 0p uc toamt i po on n, tehnet sP. C A tdraatnas fcoorl mu me rn si sa nf idt ot hne nj u satp p2 l7i e4d StDoObHo tthr atihnoi ns ge ar ensdu l tt ihneg 2 7940 S Dc oOmH p ot enset n tdsa t af ocro l ubmont hs . Tahr ee cCoDnCc ahteeanl taht emd ebaascukr teot trha ei nai ng eg , agne dn dt ee sr t, rf ae ca et u, ar en sd. Ff eoaltl ouwr ei n gd a taalsl e tt r ai sn s f2o0r0m, 5a8t i3o n sr ,o wt hs e a nt rda i n1i3n3g croowlusmannsda1n3d3tchoelutemstnsfe. ature dataset is 66,861 4.4 Modeling Ap rt etphri so cpeosisnet d, t ahne dd sapt al i ti si na tl roetardayi nfiunlgl ya n d t e s t dMaot da se el itns gf oArpMp roodaecl hi n2g . AGpr ipdrSoeaacrhc h1 CaVn di s t h e n uh ys epde rtpoairnai tmi aelti zeer satsoelroi oe sp ot hf rmooudgehl st oa ni dde n t i f y tahpep or opat icmh .aTl ambol ed e4l. 1f obreel oa wc h l mi s tosdaellli nmgo d e l s a n d hi ny pc oe nr ps iadrearma tei toenr st hc aotntsai dk ee rae dr a. nAdl loomf tsht ea tme o d e l s pr aanr admo me t_esrt aatree=i4n2i tti oa l ai zl el odwwfiot rh c o n s i s t e n t , rTeapbrloed4u.1cible results. All Models and Hyperparameters Tested in our Modeling, along with if We Set a Random State for Each
Hyperpara meters Considered
Random State?
Model
Linear Regression N/A
No
a1l0p.0h]a: [0.1, 1.0, Yes, 42 a1l0p.0h]a: [0.1, 1.0, Yes, 42
Ridge Lasso
ElasticNet a 1 l 0 p .0 h ] a , : l [ 1 0 _ . r 1 a , t 1 io .0 : , [0.1, 0.5, 0.9] EBxotorsetmineg Gradient R(XeGgBreRsesgorressor) n[ 5_ 0e s, t1i 0m0a, t2o0r s0:] , l[e0a. 0r n1 ,i n0g. 1_ r] ,a t e : m20a,x3_0d]epth: [10,
Yes, 42 Yes, 42 Yes, 42
LBiogohstt Gi nrga dMi ea nc ht i n e R( Lei gg rhet sGsBo Mr Regressor),
n[ 5_ 0e s, t1i 0m0a, t2o0r s0:] , l[e0a. 0r n1 ,i n0g. 1_ r] ,a t e : m20a,x3_0d]epth: [10,
SRuepgrpeosrstioVnec(tSoVrR) C: [0.1, 1, 10], e0p.1s,i1lo]n: [0.01, No Fc oolruema nc hi nmToadbel le t 4y .p1e) ,(al ilsl tceodmi nb itnhaet ifounr tshoefs t l e f t ht eyspt ee dr puasrianmg et ht eer tor apitni oi nngs df oart at sheattsmf oordbeol twhe r e M2 . oTdoe dl i entge rAmp pi nr eo at hc he b1 easnt dh yMpoedr eplai nr agmAeptpe rr os af cohr ea ap cphr oma oc hd)e, l f ti vy ep-ef o( lwd i ct hr oi ns se-avcahl i dmaot idoenl i wn ga s u s e d we vi at hl uraot oi ot nmme ae nt r si cq. uRaMr eSdE ecrarl oc ur l(aRt eMsStEh)e a s t h e
109
Made with FlippingBook - Online Brochure Maker