Problem
Given a set of 1-axis accelerometer measurements collected at an interval of 0.05 sec. In total, there are 10,000 measurements.
1. Calculate the dimension of the feature vector and the total number of segmentation if the time series data is segmented for every 1 sec.
2. Assuming the segmentations are to be categorized into up, right, down and left directions, explain the type of supervised learning task and the predictive model in terms of its input and output.
3. Explain the appropriate loss function to evaluate the performance of the model.