Snore Detection

The current Repository demonstrates how to detect Snoring using Ballistocardiography readings.

Workflow

1. Exploratory Data Analysis:

• Data analysis for Trends and Patterns from a Descriptive perspective.

2. Statistical Analysis:

• 5 Point Summary: Stating the Mean, Median, Min, Mix and Standard Deviation of the Sample.
• Feature Engineering: Created several features for future analysis based on stated Assumptions.
• Descriptive Analysis:
  • I study the Data Visually, in order to generate Hypothesis.
• Hypothesis Testing:
  • Based on the Hypothesis generated, I perform several statistical tests to reject/do not reject the Hypothesis.

3. Snore Detection:

• Spindle Analysis:
  • Spindle detection is performed from an 'Information' perspective.
• Loss Based Approach:
  • Detection is performed with respect to pre-set threshold.

References:

EEG sleep spindle detection using continuous wavelet transformation
Automatic sleep spindle detection
Heart Rate Measurement
Numerical Python Book