Sprint planning

[zeydabadi]

All,

Please write in the comments what is your task for the next meeting (please add your full name as well).

[960ptou]

Weijie Chen

• Working with EDA group to merge notebooks into one colab file (https://colab.research.google.com/drive/1w_c9n2dchl2SZVbIyR_XB_wTQQY99KG- )
• dataset class for data analysis and used as a ml dataset w/ pytorch (https://github.com/sharinkan/HeartDiseasePrediction/blob/eda-group/scripts/dataloader.py)
• (started on) checkout how to extract noise from audio files.

[inari-mike]

Mike

• Created scripts for using lazy loading strategy in using patient data (https://github.com/sharinkan/HeartDiseasePrediction/tree/working/scripts)
• Revised pre-processing program for Mel-Spectrum transform (https://colab.research.google.com/drive/1OK3_cdHslrnonEai7rdY2b0MR8tK8dsP?usp=sharing#scrollTo=5LPntMAqYYcl)
• Attempting conventional methods for data cleaning

[kkziao]

Ziao

Revised Short Time Fourier Transform parameter configuration.

Add a script to extract information from all txt files and form a pandas dataframe for processing.

Working with Preprocessing group to integrate all codes into one notebook.

Tried to use OpenSmile as feature extractor.

[zeydabadi]

Where are we with consolidating the features extracted so far? Is anyone working on forming tabular data that contains these features and the associated label for each instance?

[kkziao]

Where are we with consolidating the features extracted so far? Is anyone working on forming tabular data that contains these features and the associated label for each instance?

Hi Professor Zeydabadinezhad,

Yes. We currently are trying to use openSMILE to get some features out from the audio data, and we might want to try some simple Machine Learning models with the data to form our baseline models.

Thanks, Ziao You