Table 3 Numerical evaluation of the suggested movement recognition model over numerous techniques.

From: Development of weighted residual RNN model with hybrid heuristic algorithm for movement recognition framework in ambient assisted living

Terms	MLSTM³³	BiLSTM³⁶	DBN⁴³	WRRNN	HRS-COA-WRRNN
Linear activation function
Accuracy	88.65201	89.53683	90.59341	92.29874	94.14164
Sensitivity	88.93773	89.64591	90.52503	92.21001	94.06593
Specificity	88.63161	89.52904	90.59829	92.30508	94.14704
Precision	35.84822	37.94707	40.7497	46.11908	53.44433
FPR	11.36839	10.47096	9.401709	7.694924	5.852957
FNR	11.06227	10.35409	9.474969	7.789988	5.934066
NPV	88.63161	89.52904	90.59829	92.30508	94.14704
FDR	64.15178	62.05293	59.2503	53.88092	46.55567
F1-score	51.09965	53.32268	56.20073	61.48585	68.16191
MCC	0.520786	0.54218	0.569702	0.620251	0.683736
ReLU activation function
Accuracy	88.6936	89.64983	90.46561	92.25782	94.18278
Sensitivity	88.67244	89.52381	90.56277	92.03463	94.35786
Specificity	88.69511	89.65883	90.45867	92.27376	94.17027
Precision	35.90837	38.20903	40.4043	45.97088	53.62034
FPR	11.30489	10.34117	9.541332	7.726242	5.829726
FNR	11.32756	10.47619	9.437229	7.965368	5.642136
NPV	88.69511	89.65883	90.45867	92.27376	94.17027
FDR	64.09163	61.79097	59.5957	54.02912	46.37966
F1-score	51.11675	53.55894	55.87856	61.31513	68.3817
MCC	0.520404	0.544054	0.566894	0.61839	0.686235
Tanh activation function
Accuracy	88.70194	89.60494	90.49735	92.28689	94.24456
Sensitivity	88.4127	89.29453	90.40564	92.16931	94.25044
Specificity	88.7226	89.62711	90.50391	92.29529	94.24414
Precision	35.89689	38.07626	40.47694	46.07653	53.90901
FPR	11.2774	10.37289	9.496095	7.704712	5.755858
FNR	11.5873	10.70547	9.594356	7.830688	5.749559
NPV	88.7226	89.62711	90.50391	92.29529	94.24414
FDR	64.10311	61.92374	59.52306	53.92347	46.09099
F1-score	51.06188	53.38746	55.91797	61.43898	68.58756
Sigmoid activation function
Accuracy	88.6576	89.51776	90.56085	92.42479	94.14059
Sensitivity	89.04762	89.59184	90.63492	92.4263	94.30839
Specificity	88.62974	89.51247	90.55556	92.42468	94.1286
Precision	35.87284	37.89565	40.66952	46.56689	53.43011
FPR	11.37026	10.48753	9.444444	7.575316	5.871396
FNR	10.95238	10.40816	9.365079	7.573696	5.69161
NPV	88.62974	89.51247	90.55556	92.42468	94.1286
FDR	64.12716	62.10435	59.33048	53.43311	46.56989
F1-score	51.1428	53.26233	56.14553	61.93117	68.21388
MCC	0.521397	0.54153	0.569424	0.62465	0.684629
Softmax activation function
Accuracy	88.68571	89.67831	90.53968	92.35344	94.21799
Sensitivity	88.79365	89.61905	90.28571	92.4127	93.65079
Specificity	88.678	89.68254	90.55782	92.34921	94.2585
Precision	35.90501	38.28835	40.58219	46.31663	53.81248
FPR	11.322	10.31746	9.442177	7.650794	5.741497
FNR	11.20635	10.38095	9.714286	7.587302	6.349206
NPV	88.678	89.68254	90.55782	92.34921	94.2585
FDR	64.09499	61.71165	59.41781	53.68337	46.18752
F1-score	51.13346	53.6539	55.99527	61.70641	68.35032

Back to article page

Table 3 Numerical evaluation of the suggested movement recognition model over numerous techniques.

Search

Quick links