Skip to content

Latest commit

 

History

History
10 lines (8 loc) · 856 Bytes

File metadata and controls

10 lines (8 loc) · 856 Bytes

ANN_Implementation

Note that please download the dataset: train-images-idx3-ubyte.gz(http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz) train-labels-idx1-ubyte.gz(http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz) from MNIST (http://yann.lecun.com/exdb/mnist/). Before runing the program, these two file also should be unzipped and placed in /6205Project-ANNs

In this project, a small subsection of object recognition—digit recognition will be implemented, based on the MNIST database of handwritten digits. It includes building a neural network structure, training the model and accuracy evaluating.
Please download and run it locally.In console you would see the change of the loss function and final predication accuracy. For now the best result we got is 988/1000.For detail explain of this project please read the report.