MLEARN Project

Data

HAR motion-sense dataset hosted on kaggle

This dataset includes time-series data generated by accelerometer and gyroscope sensors (attitude, gravity, userAcceleration, and rotationRate). It is collected with an iPhone 6s kept in the participant’s front pocket using SensingKit which collects information from Core Motion framework on iOS devices. A total of 24 participants in a range of gender, age, weight, and height performed 6 activities in 15 trials in the same environment and conditions: downstairs, upstairs, walking, jogging, sitting, and standing. With this dataset, we aim to look for personal attributes fingerprints in time-series of sensor data, i.e. attribute-specific patterns that can be used to infer gender or personality of the data subjects in addition to their activities.

Analysis

• Correlations
• Metadata (classes, gender, age, ...)
• Fourier transform
• Wavelet transform

Feature extraction

• Fourier transform
• Wavelet transform
• Dynamic time warping
• Global alignment kernel (SVM)

Feature selection

??

Classification

Baseline

Feature extraction methods will be compared using this classification method.

Possible candidates are naive bayes, logistic regresion, ...

Advanced methods

SVM, kNN, classification trees, MLP, ...

Ensemble methods

Classification rule based on best classificators.

Boosting algorithms (XGBoost, LightGBM, ...).

Neural networks

CNNs or RNNs for classification without feature extraction.