Table 1.

Summary of classifiers which integrate deep network components into non-neural network classifiers.

	Methods	Classifiers
Architecture	Feedforward	ANT (Tanno et al. 2019), DNDT (Yang et al. 2018), DBN (Hinton et al. 2006), Deep PCA (Liong et al. 2013), DF (Zhou and Feng 2017), DPG (Damianou 2015), R2-SVM (Vinyals et al. 2012), D-SVM (Abdullah et al. 2009), DTA-LS-SVM (Wang et al. 2019a), SFDT (Frosst and Hinton 2017)
	Autoencoder	DKF (Krishnan et al. 2015), eForest (Feng and Zhou 2018), ML-SVM (Wiering and Schomaker 2014)
	Siamese Model	SDF (Utkin and Ryabinin 2018)
	Generative Adversarial Model	GAF (Zuo et al. 2018)
Optimization	Gradient Decent	DNDF (Kontschieder et al. 2015), mGBDT (Feng et al. 2018), ML-SVM (Wiering and Schomaker 2014), TAO (Carreira-Perpiñán and Tavallali 2018)
Regularization	Parameter Norm Penalty	NLP-SVM (Fung and Mangasarian 2004), GRRF (Deng and Runger 2013), RRF (Deng and Runger 2012) , SCAD-SVM (Zhang et al. 2006)
	Dropout	DART (Rashmi and Gilad-Bachrach 2015)
Hybrid Model		CondCNN Ioannou et al. (2016), DBT (Zoran et al. 2017), DCCA (Andrew et al. 2013), DNDF (Kontschieder et al. 2015), LMM (Zhao et al. 2018) , DNN+SVM: (Tang 2013) (Niu and Suen 2012) (Zareapoor et al. 2018) (Nagi et al. 2012) (Bellili et al. 2001) (Azevedo and Zanchettin 2011) NPL: (Snell et al. 2017)(Mettes et al. 2019)(Li et al. 2018) (Chen et al. 2019)