Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
David Dovodov
@ddovod
and maybe these values are not bound of output network signals, please correct me if I'm wrong
Jan Nils Ferner
@SirRade
no no, you are right
they are only weights, not activation
our activation function is tanh
I linked the wrong code
What I'm trying to say is that we need to have a way of dynamically changing the activation function
maybe you want more precision, use sigmoid.
maybe you want more speed, use ReLU.
maybe you don't have negative values, use unsigned sigmoid.
etc.
I need to create a base class functor IActivationFunction
with properties minActivation, maxActivation
so we don't have to hardcode
but until then you can set the code that we discussed up as a pull request
I will accept it
David Dovodov
@ddovod
yes, maybe it make sense
I think IActivationFunction interface is not necessary, it's possible to use any c++ object with overloaded operator ()
and create some set of heavily used activation functions in some namespace or so
Jan Nils Ferner
@SirRade
It is indeed not necesary, but the clean way to do it. The code has to follow the SOLID rules of design
this interface also allows us to easily demand Type::activation_t GetMinActivation() const and Type::activation_t GetMaxActivation() const functions
and, of course, the operator() (this is what I meant with the word "functor")
David Dovodov
@ddovod
oh yes, it also make sense
Jan Nils Ferner
@SirRade
I will implement it in the evening after ImageMagick/ImageMagick#313
Would you later like to benchmark the iris classification time again?
David Dovodov
@ddovod
yes, of course, write here please when you finish, I'll test it
Jan Nils Ferner
@SirRade
I added the changes, however I had to put in a test (supervised_learning.hpp lines 58 - 65) because the accuracy never improved enough to make the error 0. I suspect there's an error somewere hidden in an algorithm that causes this.
I'm afraid you won't see any improvements in the iris test since the changes
at least not until we found the underlying error
Jan Nils Ferner
@SirRade
btw, I didn't square the fitness because that is only mantainable with very small amounts of outputs like XOR. Maybe at some other point in time we will add optimizations for different training sizes?
Alexey
@Dopamine1
Hi guys !
I'm new here. Could someone help me with building Hippocrates on Linux?
I haven't found installation steps in readme file
Jeremy Stucki
@STJEREM
Run 'mkdir build && cd build'
Then 'cmake .. && make'
Alexey
@Dopamine1
unfortunately, this is what pops up
`
rootuser@debian:~/Desktop/Hippocrates-development$ mkdir build && cd build
rootuser@debian:~/Desktop/Hippocrates-development/build$ cmake .. && make
-- The C compiler identification is GNU 4.9.2
-- The CXX compiler identification is GNU 4.9.2
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Looking for include file pthread.h
-- Looking for include file pthread.h - found
-- Looking for pthread_create
-- Looking for pthread_create - not found
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - found
-- Found Threads: TRUE
-- Configuring done
CMake Warning (dev) at CMakeLists.txt:19 (add_library):
Policy CMP0028 is not set: Double colon in target name means ALIAS or
IMPORTED target. Run "cmake --help-policy CMP0028" for policy details.
Use the cmake_policy command to set the policy and suppress this warning.

Target "Hippocrates" links to target "Threads::Threads" but the target was
not found. Perhaps a find_package() call is missing for an IMPORTED
target, or an ALIAS target is missing?
This warning is for project developers. Use -Wno-dev to suppress it.

CMake Warning (dev) at Tests/xor_two_bits/CMakeLists.txt:3 (add_executable):
Policy CMP0028 is not set: Double colon in target name means ALIAS or
IMPORTED target. Run "cmake --help-policy CMP0028" for policy details.
Use the cmake_policy command to set the policy and suppress this warning.

Target "xor_two_bits" links to target "Threads::Threads" but the target was
not found. Perhaps a find_package() call is missing for an IMPORTED
target, or an ALIAS target is missing?
This warning is for project developers. Use -Wno-dev to suppress it.

CMake Warning (dev) at Tests/xor_three_bits/CMakeLists.txt:3 (add_executable):
Policy CMP0028 is not set: Double colon in target name means ALIAS or
IMPORTED target. Run "cmake --help-policy CMP0028" for policy details.
Use the cmake_policy command to set the policy and suppress this warning.

Target "xor_three_bits" links to target "Threads::Threads" but the target
was not found. Perhaps a find_package() call is missing for an IMPORTED
target, or an ALIAS target is missing?
This warning is for project developers. Use -Wno-dev to suppress it.

CMake Warning (dev) at Tests/EvenNumbers/CMakeLists.txt:3 (add_executable):
Policy CMP0028 is not set: Double colon in target name means ALIAS or
IMPORTED target. Run "cmake --help-policy CMP0028" for policy details.
Use the cmake_policy command to set the policy and suppress this warning.

Target "EvenNumbers" links to target "Threads::Threads" but the target was
not found. Perhaps a find_package() call is missing for an IMPORTED
target, or an ALIAS target is missing?
This warning is for project developers. Use -Wno-dev to suppress it.

-- Generating done
-- Build files have been written to: /home/rootuser/Desktop/Hippocrates-development/build
Scanning dependencies of target Core
[ 6%] Building CXX object Core/CMakeFiles/Core.dir/Sources/Implementations/utility/logger.cpp.o
c++: error: unrecognized command line option ‘-std=c++1z’
Core/CMakeFiles/Core.dir/build.make:54: recipe for target 'Core/CMakeFiles/Core.dir/Sources/Implementations/utility/logger.cpp.o' failed
make[2]: *** [Core/CMakeFiles/Core.dir/Sources/Implementations/utility/logger.cpp.o] Error 1
CMakeFiles/Makefile2:110: recipe for target 'Core/CMakeFiles/Core.dir/all' failed
make[1]: *** [Core/CMakeFiles/Core.dir/all] Error 2
Makefile:86: recipe for target 'all' failed
make: *** [all] Error 2
`
could you please tell me what should I do?
Google doesn't give useful results :(
Jeremy Stucki
@STJEREM
You need a newer version of gcc
At least gcc 6.0 is required
(I should probably add these things to the readme)
Alexey
@Dopamine1
oh my god. It was so challenging to update gcc on debian!
thank you very much. it works now ! :)
Alexey
@Dopamine1
Hi everyone!
Could you please tell me the way how can I crossover or mutate two Trained::NeuralNetwork's ?
And get their "children"
Alexey
@Dopamine1
I thought that I should work with theirs genes, but I didn't find appropriate method for this in Genome class
Jeremy Stucki
@STJEREM
Ping @SirRade
Alexey
@Dopamine1
How can I ping @SirRade?
Jeremy Stucki
@STJEREM
You just did :)
Jan Nils Ferner
@SirRade
Hello Alexey!
I am currently working hard with the team on the documentation of the image recognition module of hippocrates.
This has to be finished pretty soon. I will be happy to help you in the coming days.
I trust that you understand and thank you for your patience.
Alexey
@Dopamine1
Sounds amazing! Hippocrates getting better and better.
I have something else to do, so I am ready for some waiting.
Hope you'll finish it soon :)