Journal of Big Data

Table 7 Experimental results for different hidden layers configurations

From: A comparative dimensionality reduction study in telecom customer segmentation using deep learning and PCA

Hidden layers	Training loss (MSE)	Validation loss (MSE)
[100,50,100]	3.7788e−04	3.8346e−04
[100,30,100]	4.3803e−04	4.4357e−04
[100,20,100]	5.5041e−04	5.5685e−04
[180,100,20,100,180]	4.7450e−04	4.7450e−04
[180,100,30,100,180]	3.9926e−04	4.0332e−04
[180,100,50,100,180]	3.2366e−04	3.3278e−04

Back to article page