Coming from the PyTorch Vision ecosystem where there is a robust API for transformations, I am trying to find the analog in the Flux.jl ecosystem. I saw in the Flux docs that it mentions Augmenter.jl, is this the defacto package to do image transformations or is there a built in Flux way of doing it?
Related
I am trying to understand if the sf package in R operates at the local (desktop) level or if it uses API to transmit information online. The documentation mentions the use of API to pull in algorithms but it is unclear to me what that involves.
The package operates locally; once compiled it is perfectly capable of being run in firewalled contexts if that is your question.
I use R, but lately are trying to implement Model Predictive Control as applied in Control Engineering on my models, but it seems that this area in R is non-existent as compared to Matlab where it is quite easy to do System Identification and create transfer functions which can be deployed inside the Model Predictive Control module. Anybody knows where to look, or which packages to use for MPC and transfer functions in R to build MPC models?
There are a growing number of Python packages for control engineering. One option is to use the reticulate R package to interface to Python functions. Here are some Control Engineering packages in Python:
SciPy.signal for signal processing and system modeling
SymPy for Laplace transforms and differential equation analytic solutions
Control Systems Library
Chemical Process Control from Jeff Kantor, Notre Dame
Process Dynamics and Control in Python at BYU (my course)
Advanced Control and Machine Learning in Python at BYU (also my course)
Here is an example of running MPC in Python on the Temperature Control Lab.
Is there a pmml conversion for the bnlearn package in r, so that bayesian networks can be accessed as an API over the web?
I have seen pmml conversion for some models like decision trees and linear regression, see here, but I haven't seen any for bnlearn.
Also, is there any other service which can be used as an API for bayesian network models built in R?
I haven't tried this personally but the package r2pmml from GitHub (it is not available from CRAN) may be an option for this. You would then you "Augustus" in Python.
I'm trying to use the google-analytics framework to create predictive analysis tools. For example I would like to cluster my webpage visitors, etc.
In general, is there any list of machine learning algorithms implemented by this framework? for example: regression, clustering, classification, feature selection, etc.
Thank you for any help
Depending upon your language of choice, you might want to export your Google Analytics Metrics to flat files or a database and then start experimenting with ML models. Two popular languages with stable ML Implementations are Python and R. R's caret package includes tools for building a predictive model pipeline. Python's scikit-learn also contains implementations of all major classes of ML algorithms.
When you say GA framework I'll assume you're referring to the set of Google Analytics APIs listed here. The framework by itself doesn't provide machine learning capabilities. It merely provides access to the processed and aggregated GA data stored in Google's servers. You can use the API and feed the data to a machine learning application/system/program that does all of the stuff you mentioned.
I am searching for a neural network sample code in OpenCL, that I might optimize using GPU kernels. Please help me as I am a beginner in OpenCL.
I've started writing a deep convolutional neural network library for OpenCL at https://github.com/hughperkins/ClConvolve/tree/master . So far, it supports:
convolutional layers
max-pooling
softmax
random translations layer
random patches layer
normalization layer
multinet (aka 'multi-column', per http://arxiv.org/pdf/1202.2745.pdf )
fully-connected
... running on the GPU, using OpenCL. You can specify the network architecture on the commandline, like 100C5-MP2-100C5-MP2-100C4-MP2-300N-100N-6N.
Edit: can get 99.55% test accuracy on MNIST now :-)
You could look into this OpenCL enhanced version of FANN library for fast neural networks.
I am sure it will need some optimization :-)
I also plan to optimize THIS lib.
http://leenissen.dk/fann/wp/2011/02/opencl-enhanced-version-of-fann/