this repository is for testing some of the functionality of tensorflow
Created by Carl R. Knox on Mon Nov 9 10:43:08 MST 2015
- shows construction of a basic graph
- shows execution of ops in a graph (e.g. matrix multiplication op-type here)
- shows launching of a graph in a session
- shows use of tensorflow.device to utilize specific cores for ops
- shows creation of a scalar variable and operations on it as a tensor, using tensorflow
- shows how to fetch multiple outputs of ops so that all ops are run once, not once per requested tensor.
- shows how to patch tensors directly into any op in a graph
- shows how to utilize tensorflow to train with simple gradient descent optimization on MNIST data (ref. below).
- notes forthcoming
- notes forthcoming
- each directory contains a target script, usually named similarly to the directory name; each directory is self- contained in that way and can be run individually.
- the module scripts are interpreted using a base tensorflow virtualenv, installation instructions here
- some IDE settings are included in this repo. if you don't want them then just delete the .idea directory from project root after cloning the project.
- MNIST : Mixed National Institute of Standards and Technology
- tensorflow tutorials
- MNIST Data
- MNIST Visualization
- Principal Component Analysis
- Softmax Function Description
- Cross-Entropy
- Backpropagation Algorithm
- list of tensorflow optimization algorithms
- Rectified Linear Units (neural networks)
- Convolutional Neural Networks
- Logit Models for Binary Data
- Logit
- Modeling one neuron : neural networks