# time-series-classification

**Repository Path**: jevoncode/time-series-classification

## Basic Information

- **Project Name**: time-series-classification
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Apache-2.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2021-10-11
- **Last Updated**: 2021-10-11

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Time Series Classification

Install tsfresh (`pip install tsfresh`). 

Edit `config.py` for the dataset you want to handle. By default, it's _Wafer_. Two more are provided in the `data\` directory: _Ford A_ and _Ford B_. You can use any dataset from the [UEA & UCR Time Series Classification Repository](http://www.timeseriesclassification.com/dataset.php).

When ready, run

1. `extract_features.py`
2. `select_features.py`
3. `train_and_evaluate.py`

Step one takes some time, so you can skip it - each dataset directory already contains extracted features.

The code uses Python 2 (print statements).