1 FRESA.CAD Regresion Benchmark

1.1 OAI Joint Space Width KL Modelling Data Set

load(file = "OAIJSW_BASELINE.RDATA")


ExperimentName <- "OAIJSW_KL"
bswimsReps <- 20;
theData <- baseline;
theOutcome <- "BLKL";
reps <- 20;
fraction <- 0.2;

FRESAFileName <- paste(ExperimentName,"FRESAMethod.RDATA",sep = "_")
CVFileName <- paste(ExperimentName,"CVMethod.RDATA",sep = "_")
sampleTrain <- sample(nrow(theData),nrow(theData)*0.2)
sampleData <- theData[sampleTrain,]
testdata <- theData[-sampleTrain,]

1.2 Benchmark


theData$AGE <- as.numeric(theData$AGE)
bmodel <- BSWiMS.model(formula = paste(theOutcome," ~ 1"),data = sampleData,NumberofRepeats = bswimsReps )

save(bmodel,file = FRESAFileName)
load(file = FRESAFileName)

par(mfrow = c(2,2));
cp <- OrdinalBenchmark(theData,theOutcome,reps,fraction)

par(mfrow = c(1,1));

save(cp,file = CVFileName)
load(file = CVFileName)

1.3 Results

1.3.1 Model Selection Results


hm <- heatMaps(Outcome = "Outcome",data = cp$testPredictions,title = "Heat Map",Scale = TRUE,hCluster = "col",cexRow = 0.25,cexCol = 0.75,srtCol = 45)


#The Times
pander::pander(cp$cpuElapsedTimes)

BSWiMS	RF	RPART	LASSO	SVM	KNN	ENS
26.25	0.976	0.0955	29.16	0.142	0.0535	56.68


par(mfrow = c(2,1));
pr <- plot(cp,prefix = paste(ExperimentName,":"));

par(mfrow = c(1,1));

pander::pander(pr$metrics,caption = ExperimentName,round = 3)

OAIJSW_KL
	RF	ENS	SVM	KNN	LASSO	RPART	BSWiMS
KAPPA	0.643	0.63	0.622	0.617	0.614	0.604	0.602
BMAE	0.522	0.526	0.541	0.546	0.547	0.565	0.563
Kendall	0.58	0.569	0.561	0.567	0.553	0.545	0.548
Bias	-0.145	-0.221	-0.182	-0.368	-0.144	-0.253	-0.285
ACC	0.473	0.467	0.472	0.438	0.474	0.418	0.452
SEN	0.572	0.575	0.572	0.555	0.566	0.551	0.563
AUC	0.723	0.723	0.721	0.711	0.719	0.708	0.714

pander::pander(pr$metrics_filter,caption = ExperimentName,round = 3)

OAIJSW_KL
	BSWiMS	RPART	LASSO	mRMR	F.Test	Kendall	RF.ref
KAPPA	0.621	0.615	0.615	0.622	0.622	0.618	0.617
BMAE	0.544	0.549	0.545	0.541	0.545	0.549	0.556
Kendall	0.566	0.568	0.565	0.561	0.564	0.561	0.568
ACC	0.466	0.46	0.467	0.472	0.468	0.47	0.445
SEN	0.558	0.558	0.556	0.56	0.559	0.56	0.551
AUC	0.712	0.713	0.71	0.714	0.713	0.713	0.708

1.3.2 Radar Plots

op <- par(no.readonly = TRUE)

par(mfrow = c(1,2),xpd = TRUE,pty = "s",mar = c(1,1,1,1))

mNames <- names(cp$cpuElapsedTimes)

classRanks <- c(pr$minMaxMetrics$BMAE[1],pr$minMaxMetrics$KAPPA[2],pr$minMaxMetrics$Kendall[2],pr$minMaxMetrics$ACC[2],pr$minMaxMetrics$SEN[2],pr$minMaxMetrics$AUC[2],min(cp$cpuElapsedTimes))
classRanks <- rbind(classRanks,c(pr$minMaxMetrics$BMAE[2],0,0,0,0,0,max(cp$cpuElapsedTimes)))
classRanks <- as.data.frame(rbind(classRanks,cbind(t(pr$metrics[c("BMAE","KAPPA","Kendall","ACC","SEN","AUC"),mNames]),cp$cpuElapsedTimes)))
colnames(classRanks) <- c("BMAE","KAPPA","Kendall","ACC","SEN","AUC","CPU")

classRanks$BMAE <- -classRanks$BMAE
classRanks$CPU <- -classRanks$CPU

colors_border = c( rgb(1.0,0.0,0.0,1.0), rgb(0.0,1.0,0.0,1.0) , rgb(0.0,0.0,1.0,1.0), rgb(0.2,0.2,0.0,1.0), rgb(0.0,1.0,1.0,1.0), rgb(1.0,0.0,1.0,1.0), rgb(0.0,0.0,0.0,1.0), rgb(1.0,1.0,0.0,1.0) )
colors_in = c( rgb(1.0,0.0,0.0,0.05), rgb(0.0,1.0,0.0,0.05) , rgb(0.0,0.0,1.0,0.05),rgb(1.0,1.0,0.0,0.05), rgb(0.0,1.0,1.0,0.05) , rgb(1.0,0.0,1.0,0.05), rgb(0.0,0.0,0.0,0.05), rgb(1.0,1.0,0.0,0.05) )
fmsb::radarchart(classRanks,axistype = 0,maxmin = T,pcol = colors_border,pfcol = colors_in,plwd = c(6,2,2,2,2,2,2),plty = 1, cglcol = "grey", cglty = 1,axislabcol = "black",cglwd = 0.8, vlcex  = 0.5 ,title = "Prediction Model")

legend("topleft",legend = rownames(classRanks[-c(1,2),]),bty = "n",pch = 20,col = colors_in,text.col = colors_border,cex = 0.5,pt.cex = 2)


filnames <- c("BSWiMS","LASSO","RF.ref","F.Test","Kendall","mRMR")

filterRanks <- c(pr$minMaxMetrics$BMAE[1],pr$minMaxMetrics$KAPPA[2],pr$minMaxMetrics$Kendall[2],pr$minMaxMetrics$ACC[2],pr$minMaxMetrics$SEN[2],pr$minMaxMetrics$AUC[2],max(cp$jaccard),min(cp$featsize));

filterRanks <- rbind(filterRanks,c(pr$minMaxMetrics$BMAE[2],0,0,0,0,0,min(cp$jaccard),max(cp$featsize)));

filterRanks <- as.data.frame(rbind(filterRanks,cbind(t(pr$metrics_filter[c("BMAE","KAPPA","Kendall","ACC","SEN","AUC"),filnames]),cp$jaccard[filnames],cp$featsize[filnames])));
colnames(filterRanks) <- c("BMAE","KAPPA","Kendall","ACC","SEN","AUC","Jaccard","SIZE")
filterRanks$BMAE <- -filterRanks$BMAE
filterRanks$SIZE <- -filterRanks$SIZE

colors_border = c( rgb(1.0,0.0,0.0,1.0), rgb(0.0,1.0,0.0,1.0) , rgb(0.0,0.0,1.0,1.0), rgb(0.2,0.2,0.0,1.0), rgb(0.0,1.0,1.0,1.0), rgb(1.0,0.0,1.0,1.0), rgb(0.0,0.0,0.0,1.0) )
colors_in = c( rgb(1.0,0.0,0.0,0.05), rgb(0.0,1.0,0.0,0.05) , rgb(0.0,0.0,1.0,0.05),rgb(1.0,1.0,0.0,0.05), rgb(0.0,1.0,1.0,0.05) , rgb(1.0,0.0,1.0,0.05), rgb(0.0,0.0,0.0,0.05) )
fmsb::radarchart(filterRanks,axistype = 0,maxmin = T,pcol = colors_border,pfcol = colors_in,plwd = c(6,2,2,2,2,2,2),plty = 1, cglcol = "grey", cglty = 1,axislabcol = "black",cglwd = 0.8, vlcex  = 0.6,title = "Filter Method" )


legend("topleft",legend = rownames(filterRanks[-c(1,2),]),bty = "n",pch = 20,col = colors_in,text.col = colors_border,cex = 0.5,pt.cex = 2)


par(mfrow = c(1,1))
par(op)

1.3.3 Features Analysis


pander::pander(summary(bmodel),caption = "Model",round = 3)

coefficients:

Table continues below
	Estimate	lower	OR	upper	u.Accuracy
STDJSW	0.1578	1.145	1.171	1.197	0.5999
MtoLDiff	0.1812	1.168	1.199	1.23	0.6009
CVJSW	0.2191	1.207	1.245	1.284	0.6175
stdLJSW	-0.2358	0.7579	0.79	0.8235	0.5702
LJSW725	-0.0326	0.9626	0.9679	0.9733	0.5461
lCV	0.07686	1.067	1.08	1.093	0.4513
MaxCV	0.08357	1.071	1.087	1.104	0.5749
meanLJSW	-0.1406	0.8443	0.8688	0.8941	0.5119
BMI	0.03231	1.025	1.033	1.041	0.5742
LJSW700	-0.09316	0.8903	0.911	0.9322	0.5639
LJSW750	-0.03963	0.9519	0.9611	0.9704	0.5292
JSW175	-0.03729	0.9538	0.9634	0.9731	0.59
AGE	0.01564	1.011	1.016	1.02	0.582
LateralSlope	0.07017	1.052	1.073	1.094	0.5568
DiffSlope	-0.04718	0.9394	0.9539	0.9687	0.5107
JSW200	-0.006328	0.9916	0.9937	0.9958	0.5735
WEIGHT	0.00562	1.004	1.006	1.008	0.5687
LJSW825	0.09117	1.057	1.095	1.136	0.488

Table continues below
	r.Accuracy	full.Accuracy	u.AUC	r.AUC	full.AUC
STDJSW	0.5582	0.6217	0.6295	0.5594	0.681
MtoLDiff	0.5823	0.6439	0.6293	0.588	0.6797
CVJSW	0.6555	0.6709	0.6604	0.651	0.7049
stdLJSW	0.642	0.6604	0.5593	0.673	0.7018
LJSW725	0.6157	0.6433	0.5376	0.6423	0.6835
lCV	0.5965	0.6316	0.4773	0.6476	0.6872
MaxCV	0.5876	0.6057	0.6101	0.6155	0.6512
meanLJSW	0.6266	0.6473	0.5302	0.6618	0.6813
BMI	0.644	0.6709	0.5973	0.6993	0.7049
LJSW700	0.661	0.6787	0.5703	0.6897	0.712
LJSW750	0.597	0.6168	0.5285	0.6561	0.6747
JSW175	0.5616	0.6028	0.6071	0.6228	0.6397
AGE	0.6465	0.6709	0.5854	0.6913	0.7049
LateralSlope	0.6015	0.6121	0.5477	0.6393	0.6658
DiffSlope	0.5966	0.6013	0.5012	0.6265	0.6409
JSW200	0.5668	0.5944	0.6064	0.628	0.6429
WEIGHT	0.6271	0.6439	0.5793	0.6654	0.6797
LJSW825	0.6385	0.6484	0.4981	0.6735	0.6824

	IDI	NRI	z.IDI	z.NRI	Frequency
STDJSW	0.1459	0.6163	16.43	13.24	0.2676
MtoLDiff	0.1374	0.554	15.9	11.63	0.2817
CVJSW	0.1364	0.6175	15.87	13.15	0.2817
stdLJSW	0.05271	0.48	9.093	9.966	0.2817
LJSW725	0.04506	0.3425	8.469	7.018	0.2676
lCV	0.04301	0.3859	8.406	8.008	0.08451
MaxCV	0.04164	0.3049	8.037	6.281	0.1972
meanLJSW	0.03761	0.3666	7.814	7.503	0.1127
BMI	0.03254	0.4073	7.25	8.39	0.2817
LJSW700	0.03083	0.37	7.245	7.588	0.07042
LJSW750	0.03111	0.3301	7.188	6.735	0.1127
JSW175	0.02742	0.236	6.628	4.791	0.1408
AGE	0.0261	0.2685	6.584	5.453	0.2817
LateralSlope	0.02534	0.274	6.001	5.591	0.1972
DiffSlope	0.02479	0.2613	5.944	5.303	0.1127
JSW200	0.01988	0.2164	5.769	4.38	0.02817
WEIGHT	0.01783	0.3014	5.291	6.135	0.2817
LJSW825	0.01538	0.1637	4.857	3.305	0.09859

Accuracy: 0.662
tAUC: 0.7147
sensitivity: 0.6295
specificity: 0.8
bootstrap:

coefficients:

Table continues below
	Estimate	lower	OR	upper	u.Accuracy
MtoLDiff	0.1457	1.133	1.157	1.181	0.6054
CVJSW	0.2034	1.191	1.226	1.261	0.6482
STDJSW	0.1387	1.126	1.149	1.172	0.6082
LJSW725	-0.03076	0.9648	0.9697	0.9747	0.5675
LJSW700	-0.2717	0.7268	0.7621	0.799	0.5866
meanLJSW	-0.2569	0.7361	0.7734	0.8126	0.5439
LJSW750	-0.06743	0.9233	0.9348	0.9465	0.5442
MaxCV	0.0509	1.041	1.052	1.063	0.5809
stdLJSW	-0.0395	0.9542	0.9613	0.9684	0.5558
AGE	0.01691	1.011	1.017	1.023	0.5779
lCV	0.01352	1.01	1.014	1.018	0.4695
DiffSlope	-0.05237	0.9315	0.949	0.9668	0.5252
BMI	0.0267	1.017	1.027	1.038	0.5477
LJSW825	0.01982	1.012	1.02	1.028	0.5079
LatIntersept	0.1393	1.089	1.149	1.213	0.5293
JSW225	-0.01162	0.9838	0.9884	0.9931	0.6057
JSW200	-0.04373	0.9418	0.9572	0.9729	0.601
LJSW800	0.142	1.079	1.153	1.232	0.5251
LatCurvature	-0.06396	0.9084	0.938	0.9687	0.5407
JSW275	-0.005035	0.9921	0.995	0.9978	0.6051

Table continues below
	r.Accuracy	full.Accuracy	u.AUC	r.AUC	full.AUC
MtoLDiff	0.5804	0.6723	0.6249	0.5851	0.6943
CVJSW	0.6402	0.6929	0.6679	0.6393	0.717
STDJSW	0.5903	0.6591	0.6253	0.5983	0.686
LJSW725	0.6217	0.6675	0.5636	0.6392	0.6944
LJSW700	0.6649	0.6929	0.5893	0.6855	0.717
meanLJSW	0.632	0.6723	0.5546	0.6522	0.6943
LJSW750	0.6261	0.6599	0.5457	0.6485	0.6853
MaxCV	0.5989	0.6216	0.6028	0.6196	0.6477
stdLJSW	0.6258	0.6474	0.5483	0.6492	0.6796
AGE	0.677	0.6929	0.5804	0.7038	0.717
lCV	0.6311	0.6489	0.4843	0.659	0.6829
DiffSlope	0.619	0.625	0.5264	0.6412	0.652
BMI	0.6852	0.6929	0.5576	0.7105	0.717
LJSW825	0.662	0.666	0.5189	0.6825	0.6887
LatIntersept	0.6632	0.675	0.5368	0.6833	0.6966
JSW225	0.6051	0.6253	0.6244	0.6399	0.6504
JSW200	0.637	0.6394	0.6262	0.6571	0.6634
LJSW800	0.6873	0.6929	0.5315	0.7109	0.717
LatCurvature	0.6517	0.6634	0.5495	0.6757	0.6877
JSW275	0.6687	0.6735	0.6175	0.6966	0.6969

	IDI	NRI	z.IDI	z.NRI	Frequency
MtoLDiff	0.1552	0.6581	15.66	12.8	0.2817
CVJSW	0.1312	0.6454	14.43	12.49	0.2817
STDJSW	0.107	0.5572	12.45	10.8	0.2676
LJSW725	0.06761	0.4612	9.819	8.697	0.07042
LJSW700	0.04927	0.4902	8.418	9.239	0.2817
meanLJSW	0.04804	0.4412	8.163	8.264	0.2817
LJSW750	0.04706	0.3216	8.147	5.996	0.1549
MaxCV	0.04249	0.3592	7.836	6.813	0.1549
stdLJSW	0.0453	0.4187	7.594	7.846	0.05634
AGE	0.02533	0.3268	5.933	6.041	0.2817
lCV	0.02456	0.3253	5.891	6.143	0.05634
DiffSlope	0.02005	0.2684	5.425	4.943	0.169
BMI	0.0187	0.3147	5.024	5.826	0.2817
LJSW825	0.01929	0.2003	4.929	3.68	0.02817
LatIntersept	0.01853	0.221	4.899	4.063	0.1831
JSW225	0.01897	0.2766	4.724	5.106	0.05634
JSW200	0.01854	0.3062	4.478	5.673	0.2254
LJSW800	0.01393	0.2829	4.099	5.215	0.2817
LatCurvature	0.01192	0.239	3.78	4.396	0.2676
JSW275	0.007638	0.2661	3.395	4.895	0.02817

Accuracy: 0.6891
tAUC: 0.7194
sensitivity: 0.628
specificity: 0.8108
bootstrap:

coefficients:

Table continues below
	Estimate	lower	OR	upper	u.Accuracy
STDJSW	0.3731	1.378	1.452	1.531	0.8173
MtoLDiff	0.3707	1.375	1.449	1.527	0.8226
CVJSW	0.3344	1.333	1.397	1.465	0.8742
MaxCV	0.1559	1.143	1.169	1.195	0.7846
LJSW750	-0.3353	0.682	0.7151	0.7497	0.5503
MeanJSW	-0.3172	0.6963	0.7282	0.7615	0.6691
LJSW725	-0.06029	0.9335	0.9415	0.9495	0.5715
DiffSlope	-0.1177	0.8706	0.889	0.9078	0.534
JointSlope	0.1657	1.15	1.18	1.211	0.7327
JSW200	-0.005371	0.9936	0.9946	0.9957	0.8026
meanMJSW	-0.06877	0.9222	0.9335	0.945	0.755
JSW150	-0.05334	0.9388	0.9481	0.9574	0.7874
mCV	0.1557	1.129	1.168	1.209	0.7188
lCV	-0.007077	0.9915	0.9929	0.9944	0.5318
LatSlope	0.1468	1.111	1.158	1.207	0.5975
MCMJSW	-0.1648	0.8054	0.8481	0.8931	0.7774
AGE	0.01994	1.013	1.02	1.027	0.615
TPCFDS	0.1339	1.086	1.143	1.204	0.5137
LatIntersept	0.008849	1.005	1.009	1.013	0.5345
XMJSW	-0.08443	0.8818	0.919	0.9578	0.5089

Table continues below
	r.Accuracy	full.Accuracy	u.AUC	r.AUC	full.AUC
STDJSW	0.7605	0.8981	0.7982	0.7561	0.8976
MtoLDiff	0.7568	0.8917	0.8134	0.7434	0.8874
CVJSW	0.8221	0.8968	0.8577	0.8217	0.893
MaxCV	0.7417	0.8561	0.7492	0.7344	0.8364
LJSW750	0.8182	0.9002	0.5392	0.8	0.8999
MeanJSW	0.8377	0.892	0.6693	0.8328	0.8877
LJSW725	0.8385	0.8926	0.5661	0.8317	0.892
DiffSlope	0.826	0.8561	0.5254	0.8084	0.8364
JointSlope	0.871	0.8858	0.7387	0.8597	0.88
JSW200	0.8254	0.8485	0.7845	0.8079	0.8327
meanMJSW	0.8524	0.8581	0.751	0.8273	0.8414
JSW150	0.8898	0.8926	0.7836	0.8798	0.8923
mCV	0.872	0.8917	0.6937	0.8669	0.8874
lCV	0.8519	0.8652	0.4696	0.8436	0.8623
LatSlope	0.8909	0.8968	0.5994	0.8914	0.893
MCMJSW	0.8923	0.8968	0.7885	0.8859	0.893
AGE	0.8963	0.8968	0.6255	0.8895	0.893
TPCFDS	0.8848	0.893	0.5081	0.881	0.8891
LatIntersept	0.8497	0.8555	0.5209	0.8354	0.837
XMJSW	0.8835	0.8889	0.5202	0.8785	0.8852

	IDI	NRI	z.IDI	z.NRI	Frequency
STDJSW	0.3266	1.403	27.58	38.07	0.2817
MtoLDiff	0.3263	1.366	27.32	36.01	0.2676
CVJSW	0.2802	1.469	24.94	41.94	0.2817
MaxCV	0.2225	1.113	21.54	26.17	0.1549
LJSW750	0.2262	1.237	21.38	30.36	0.1972
MeanJSW	0.1358	1.026	15.39	23.05	0.2817
LJSW725	0.1222	1.098	14.49	25.36	0.07042
DiffSlope	0.05606	0.6137	9.558	12.53	0.1549
JointSlope	0.053	0.8572	9.267	18.85	0.2676
JSW200	0.06579	0.5641	8.819	11.69	0.05634
meanMJSW	0.05177	0.5558	8.661	11.27	0.09859
JSW150	0.04323	0.9745	7.939	21.51	0.08451
mCV	0.03695	0.534	7.592	10.66	0.2676
lCV	0.03173	0.6481	7.294	13.48	0.04225
LatSlope	0.02309	0.4386	6.58	8.748	0.2817
MCMJSW	0.0274	0.7603	5.948	15.93	0.2817
AGE	0.02148	0.4313	5.632	8.513	0.2817
TPCFDS	0.0141	0.3063	5.042	5.964	0.2817
LatIntersept	0.007349	0.3108	4.312	6.12	0.05634
XMJSW	0.009117	0.1951	3.88	3.805	0.2394

Accuracy: 0.9067
tAUC: 0.9074
sensitivity: 0.9087
specificity: 0.906
bootstrap:

coefficients:

Table continues below
	Estimate	lower	OR	upper	u.Accuracy	r.Accuracy
CVJSW	0.2267	1.215	1.254	1.295	0.9542	0.04213
MaxCV	0.3666	1.37	1.443	1.519	0.919	0.6679
STDJSW	0.2605	1.248	1.298	1.349	0.9107	0.7268
MtoLDiff	0.512	1.552	1.669	1.794	0.9132	0.7859
DiffSlope	-0.6023	0.5029	0.5476	0.5961	0.6384	0.9206
MeanJSW	-0.6765	0.4621	0.5084	0.5593	0.7859	0.9132
LJSW725	-0.09308	0.8992	0.9111	0.9232	0.6994	0.9107
LJSW750	-0.1446	0.8479	0.8654	0.8832	0.6792	0.9108

Table continues below
	full.Accuracy	u.AUC	r.AUC	full.AUC	IDI	NRI
CVJSW	0.9542	0.9638	0.5	0.9638	0.888	1.858
MaxCV	0.9485	0.8892	0.6216	0.9535	0.6806	1.734
STDJSW	0.9361	0.9306	0.6768	0.9538	0.5911	1.604
MtoLDiff	0.9532	0.9432	0.7709	0.9753	0.4671	1.839
DiffSlope	0.9485	0.5836	0.8932	0.9535	0.1846	1.393
MeanJSW	0.9532	0.7709	0.9432	0.9753	0.1553	1.416
LJSW725	0.9348	0.6584	0.9306	0.9546	0.1225	1.222
LJSW750	0.9324	0.6082	0.9307	0.9529	0.1102	0.9639

	z.IDI	z.NRI	Frequency
CVJSW	128	112	0.2817
MaxCV	70.49	85.11	0.2676
STDJSW	57.23	84.58	0.1831
MtoLDiff	42.4	104	0.2817
DiffSlope	21.03	43.18	0.2676
MeanJSW	19.8	45.03	0.2817
LJSW725	16.3	34.61	0.05634
LJSW750	14.76	25.27	0.09859

Accuracy: 0.9569
tAUC: 0.9775
sensitivity: 1
specificity: 0.955
bootstrap:


topFeat <- min(ncol(bmodel$bagging$formulaNetwork),30);
shortformulaNetwork <- bmodel$bagging$formulaNetwork[1:topFeat,1:topFeat]
validf <- diag(shortformulaNetwork) > 0.1
gplots::heatmap.2(shortformulaNetwork[validf,validf],trace="none",mar = c(10,10),main = "B:SWiMS Formula Network",cexRow = 0.6,cexCol = 0.6)



rm <- rowMeans(cp$featureSelectionFrequency)
selFrequency <- cp$featureSelectionFrequency[rm > 0.10,]


gplots::heatmap.2(selFrequency,trace = "none",mar = c(10,10),main = "Features",cexRow = 0.2)

hm <- heatMaps(Outcome = theOutcome,data = theData[,c(theOutcome,rownames(selFrequency))],title = "Heat Map",Scale = TRUE,hCluster = "col",cexRow = 0.3,cexCol = 0.3,srtCol = 45)


vlist <- rownames(selFrequency)
vlist <- cbind(vlist,vlist)
univ <- univariateRankVariables(variableList = vlist,formula = paste(theOutcome,"~1"),Outcome = theOutcome,data = theData,type = "LM",rankingTest = "Ztest",uniType = "Regression")[,c("cohortMean","cohortStd","kendall.r","kendall.p")] 


cnames <- colnames(univ);
univ <- cbind(univ,rm[rownames(univ)])
colnames(univ) <- c(cnames,"Frequency")
univ <- univ[order(-univ[,5]),]
pander::pander(univ[1:topFeat,],caption = "Features",round = 4)

Features
	cohortMean	cohortStd	kendall.r	kendall.p	Frequency
MtoLDiff	1.182	2.069	0.3927	0	1
CVJSW	1.507	2.884	0.4423	0	0.9929
MCMJSW	-0.7973	1.774	-0.3315	0	0.9929
MaxCV	0.9501	3.121	0.3069	0	0.9786
STDJSW	1.097	2.08	0.3689	0	0.9714
BMI	29.26	4.83	0.1276	0	0.8714
MeanJSW	-0.3743	1.454	-0.2526	0	0.7929
AGE	61.89	9.079	0.1511	0	0.7857
LJSW700	-0.2314	1.329	-0.1714	0	0.7714
JSW150	-0.708	1.699	-0.3147	0	0.6571
mCV	0.7041	2.448	0.2305	0	0.6571
JointSlope	0.5762	2.249	0.2262	0	0.6429
JSW175	-0.7255	1.726	-0.3138	0	0.6143
WEIGHT	83.09	16.02	0.1106	0	0.6071
JSW200	-0.8022	1.895	-0.3144	0	0.5929
meanLJSW	-0.0901	1.616	-0.0605	0	0.5929
LatSlope	0.2722	1.205	0.1632	0	0.5857
meanMJSW	-0.7163	1.79	-0.3152	0	0.5643
JSW225	-0.7515	1.826	-0.3145	0	0.5571
MedIntersept	-0.7361	1.798	-0.3169	0	0.55
LJSW750	-0.1362	1.522	-0.1001	0	0.5071
stdLJSW	-0.169	1.063	-0.108	0	0.5071
LJSW825	-0.0113	1.615	-0.0108	0.3081	0.5071
LatCurvature	-0.1259	1.109	-0.0647	0	0.5071
JSW300	-0.3912	1.365	-0.2607	0	0.4929
LJSW725	-0.1821	1.448	-0.136	0	0.4714
LJSW775	-0.097	1.546	-0.0783	0	0.4714
LJSW800	-0.038	1.583	-0.0447	0	0.4571
LJSW850	-0.0037	1.676	0.0118	0.2649	0.45
LatIntersept	-0.0324	1.561	-0.0395	2e-04	0.4357

Benchmarks for Modeling OAI KL Scores

José Tamez-Peña

April 30, 2018

1 FRESA.CAD Regresion Benchmark

1.1 OAI Joint Space Width KL Modelling Data Set

1.2 Benchmark

1.3 Results

1.3.1 Model Selection Results

1.3.2 Radar Plots

1.3.3 Features Analysis