Data Set Research

[hooman007]

Find free online datasets useful for our signal processing portion of project.

Comment all your findings.

GUIDING QUESTIONS. try to answer as many as possible.

1. Are there any data sets for biometrics online? -Are they free? -If yes, what kind of data are they? (eg. csv, etc) -What bio markers are they
2. Are there any data sets for SIMILAR to biometrics data online? -what kind of data? (eg. stock market, house price, other 1-dimensional features, i.e. non-image data)

Suggested website: https://www.kaggle.com/datasets

Suggested Topics:

• mental health ( ecg, sleep cycle, appetite, weight loss, respiration, perspiration )
• other topics ( house price, stock market )
• any topic with 1-dimensional features on y-axis and time on x-axis

Suggested tags in search criteria:

• mental health
• real-time computing
• healthcare

look at kernels, other people's solutions, to learn and use their:

• data handling methods ( data reading, data augmentation, caching, etc )
• machine learning algorithms
• deep learning architectures for the purpose

[MasoudMo]

Here are some useful databases I found:

---

Original Source (Explains the Data): https://physionet.org/physiobank/database/nstdb/

.csv Files + Useful Kernel: https://www.kaggle.com/shayanfazeli/heartbeat

Comments: Very useful data loading techniques are shown in the Kernel. It also shows feature extraction from ECG signals, which is exactly what we are looking for.

Kaggle: https://www.kaggle.com/ecerulm/apneaecg

Other Useful Related Sources: https://www.physionet.org/physiobank/database/apnea-ecg/ https://www.physionet.org/physiotools/apdet/

Comments: Apnea is a breathing disorder while sleeping. Seems very similar to our application.

Kaggle: https://www.kaggle.com/dgawlik/nyse

Comments: New York Stock Exchange Data. Many useful kernels that use LSTMs to predict stock prices.

Kaggle: https://www.kaggle.com/imnikhilanand/heart-attack-prediction

Comments: Heart related data (Not properly labelled). Kernels predict chance of heart attack.

Kaggle: https://www.kaggle.com/ash316/eda-to-prediction-dietanic

Comments: Great Learning Experience for all of us before moving on to more complicated machine learning architectures.