Is this thread dead?
For all the people that are still interested in NNLP, here some interesting facts:
* On Virus Total, only the following antivirus engines generate a warning:
Cybereason Malicious4eef91
eGambit UnsafeAI_Score_99%
Malwarebytes MachineLearning/Anomalous100%
MaxSecure TrojanMalware300983susgen
SecureAge Malicious
SentinelOne (Static ML) Static AI - Suspicious PE
Webroot W32AdwareGen
And why is that? Well, the reason is that, the exe is protected by Crypto Obfuscator which is a stand-alone program that performs obfuscation and tamper prevention operations on a compiled software, so some antivirus this is a malware.
* The default Learning Rate is 9E-06 which is 0.000006, the values can go from 10 to 1000000. The default value in the interface is shown as 90/1000000.0 = 0.00009 = 9E-05
* The default Momentum is 0.31, the values can go from 10 to 1000000. The default value in the interface is shown as 25000/1000000.0 = 0.025
* The default Noise is 0.17, the values can go from 1 to 1000. The default value in the interface is shown as 250/1000.0 = 0.25
* The default Threshold Error is 0.0001, the values can go from 0 to 700000. The default value in the interface is shown as 0/100000.0 = 0
* The default Hidden Layer Neurons is 50, the values can go from 2 to 500. The default value in the interface is shown as 50
* The default Training Data Range is 50, the values can go from 3 to 20000. The default value in the interface is shown as 50
* The default Skip Training Data is 0, the values can go from 0 to 100. The default value in the interface is shown as 0
* The default Fixed Seed Input Weight is 1, the values can go from 1 to 10000. The default value in the interface is shown as 1
* The default values are those assigned to the variables but they do not reflect the ones shown in the interface.
