-
Purpose: recognition of a curve with a pair of bursts.
-
Usage: inferencer.py XLEFTMOST XRIGHTMOST
-
PARAMETERS:
XLEFTMOST bottom boundary of the examined graph's slots.
XRIGHTMOST top boundary of the examined graph's slots.
- Description:
From different, somehow distributed normalized graphs (datapoints) Hydra can determine curves with a pair of spikes (two peaks/bursts) using a deep learning architecture of a multi-layer perceptron. The number of features (slots, bins, sectors, disjoint categories) must be greater than or equal to 100 for each datapoint. Those slots form the X axis of the graph. The value of each normalized feature lies in the range from 0 to 1. These values form the graph's Y axis. The graph's X and Y in combination represent some shape (curve) (in other words, a feature vector with (XRIGHTMOST - XLEFTMOST - 1) elements).
For its task Hydra randomly generates a dataset with datapoints (feature vectors and their corresponding True(1)/False(0) labels) based on sampling from a number of probability density functions, trains a model on this dataset, and labels a new given datapoint (makes prediction) based on the model's training. Saves/Loads the new datapoint (in the NPZ format) and/or the trained model (in the TensorFlow SavedModel format) for future use, displays the processed datapoint.
- Software used during development: Python 3.8.7, Keras 2.4.3, Numpy 1.19.5, Matplotlib 3.3.4
- Call example:
inferencer.py 1 102
- Prediction examples:
| Two Bursts Curves | Other Curves |
|---|---|
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |