Time-series features

The script expects data in a folder named data inside the project. It reads every CSV file in its subfolders; each file holds accelerometer readings (X, Y and Z axes) and a column named type with the activity for each data point.

• Concatenation. All data points from the CSV files are merged into a single pandas DataFrame.
• Standardisation. A StandardScaler puts the X, Y and Z axes on a common scale for better model performance.
• Segmentation. The data is divided into fixed-length windows of accelerometer readings.
• Feature extraction. For each segment and axis: mean, standard deviation, median, minimum, maximum, range and interquartile range (IQR).
• Feature storage. Features go into a list named segments; the activity of each segment’s first element goes into labels.
• Feature DataFrame. features_df is built from segments, with labels added as the type column.

• Random Forest Classifier. An ensemble method known for its effectiveness and robustness in classification problems.
• Linear Support Vector Classifier (SVC). A linear classifier that separates data points into distinct classes, a strong candidate for activity recognition.

• Python
• pandas
• NumPy
• Matplotlib
• scikit-learn

Author: Catalin Bondari. Dublin City University coursework.