SOCR: Data Value Metric (DVM)

The SOCR DVM project provides a novel measure, called Data Value Metric (DVM), to quantify the utility and energy, or information content, of large and complex datasets. DVM can be used to determine if appending, expanding, or otherwise augmenting the data size or complexity may be beneficial in specific application domains.
Table 1: Test results by signal profile (Supervised methods & Strong signals)

Dataset	Method	Sample-size plot	Number of features plot	3D DVM Plot
MNIST	KNN	MNIST_KNN_Sample	MNIST_KNN_Feature	MNIST_KNN_3D_DVM_Surface
MNIST	Boosting	MNIST_Boosting_Sample	MNIST_Boosting_Feature	MNIST_Boosting_3D_DVM_Surface
MNIST	Linear Regression	MNIST_LinearRegression_Sample	MNIST_LinearRegression_Feature	MNIST_LinearRegression_3D_DVM_Surface
MNIST	Lasso Regression	MNIST_LassoRegression_Sample	MNIST_LassoRegression_Feature	MNIST_LassoRegression_3D_DVM_Surface
Simulated	KNN	Simulated_KNN_Sample	Simulated_KNN_feature	Simulated_KNN_3D_DVM_Surface
Simulated	Linear Regression	Simulated_LinearRegression_Sample	Simulated_LinearRegression_Feature	Simulated_LinearRegression_3D_DVM_Surface
Simulated	Lasso Regression	Simulated_LassoRegression_Sample	Simulated_LassoRegression_Feature	Simulated_LassoRegression_3D_DVM_Surface

Table 2: Test results by the signal profile (Supervised methods & Weak signals)

Dataset	Method	Sample-size plot	Number of features plot	3D DVM Plot
ALS	KNN	ALS_KNN_Sample	ALS_KNN_Feature	ALS_KNN_3D_DVM_Surface
ALS	Boosting	ALS_Boosting_Sample	ALS_Boosting_Feature	ALS_Boosting_3D_DVM_Surface
ALS	Random Forest	ALS_RandomForest_Sample	ALS_RandomForest_Feature	ALS_RandomForest_3D_DVM_Surface
ALS	Linear Regression	ALS_LinearRegression_Sample	ALS_LinearRegression_Feature	ALS_LinearRegression_3D_DVM_Surface
ALS	Lasso Regression	ALS_LassoRegression_Sample	ALS_LassoRegression_Feature	ALS_LassoRegression_3D_DVM_Surface
Simulated	KNN	Simulated_KNN_Sample	Simulated_KNN_Feature	Simulated_KNN_3D_DVM_Surface
Simulated	Boosting	Simulated_Boosting_Sample	Simulated_Boosting_Feature	Simulated_Boosting_3D_DVM_Surface
Simulated	Linear Regression	Simulated_LinearRegression_Sample	Simulated_LinearRegression_Feature	Simulated_LinearRegression_3D_DVM_Surface
Simulated	Lasso Regression	Simulated_LassoRegression_Sample	Simulated_LassoRegression_Feature	Simulated_LassoRegression_3D_DVM_Surface

Table 3: Test results by signal profile (Unsupervised methods & Strong signals)

Dataset	Method	Sample-size plot	Number of features plot	3D DVM Plot
MNIST	KMeans	MNIST_KMeans_Sample	MNIST_KMeans_Feature	MNIST_KMeans_3D_DVM_Surface
MNIST	Affinity Propagation	MNIST_AP_Sample	MNIST_AP_Feature	MNIST_AP_3D_DVM_Surface
MNIST	Agglomerative Clustering	MNIST_Agglomerative_Sample	MNIST_Agglomerative_Feature	MNIST_Agglomerative_3D_DVM_Surface
Simulated	KMeans	Simulated_KMeans_Sample	Simluated_KMeans_Feature	Simluated_KMeans_3D_DVM_Surface
Simulated	Affinity Propagation	Simulated_AP_Sample	Simulated_AP_Feature	Simluated_AP_3D_DVM_Surface

Table 4: Test results by signal profile (Unsupervised methods & Weak signals)

Dataset	Method	Sample-size plot	Number of features plot	3D DVM Plot
ALS	Affinity Propagation	ALS_AP_Sample	ALS_AP_Feature	ALS_AP_3D_DVM_Surface
ALS	Agglomerative Clustering	ALS_Agglomerative_Sample	ALS_Agglomerative_Feature	ALS_Agglomerative_3D_DVM_Surface
Simulated	Affinity Propagation	Simulated_AP_Sample	Simulated_AP_Feature	Simulated_AP_3D_DVM_Surface

Reference: Noshad, M, Choi, J, Sun, Y, Hero, A, Dinov, ID. (2021) A data value metric for quantifying information content and utility, Journal of Big Data, DOI: 10.1186/s40537-021-00446-6, in print.