Nueral Network Accuracy Predictor

Nueral Networks is a computing system made up of a number of simple, highly interconnected processing elements, which process information by their dynamic state response to external inputs.

Refer Blog for more details

EXAMPLE DATASETS

Link to the drive file

1 - Upload file

2 - Input Characteristics

3 - Get Accuracy

FORMAT OF DATASET

Since, this app does not pre process data, it must be done before feeding into the application.

Dataset must consist of numerical data.

Dataset must have "Outputs" column.

WORKING EXAMPLE

Refer to the example datasets and select Star39552_balanced.csv

This is a star classification dataset. Refer Kaggle

Test out different activation functions and hidden layer nodes.

For the above combination, accuracy is very high as 0.8785 :

But by changing the hidden nodes, accuracy can drop as low as 0.49: