Experiments
Our analysis was designed to be rigorous and consultant of real-world utilization. For each the 30-minute and 60-minute time horizons, we evaluated predictions on 100 randomly chosen stations, sampling their occupancy standing 48 instances every day (each half-hour) for a full week.
The mannequin was benchmarked towards a remarkably robust baseline: the “Maintain Present State” method. This baseline merely assumes that the variety of obtainable ports a sure variety of minutes (H) sooner or later can be precisely the identical as the present quantity.
Whereas easy, this baseline may be very onerous to beat, particularly over brief horizons. For instance, our information confirmed that on the US East Coast, by no means greater than 10% of ports change their availability state inside a 30-minute block. Since more often than not the state does not change, the only prediction — no change — is appropriate more often than not, making the duty of including predictive worth extraordinarily tough.
We targeted on two key metrics to measure the mannequin’s accuracy for predicting the precise variety of free ports: imply squared error (MSE) and imply absolute error (MAE). A ratio of MSE/MAE ≥ 1 free port measures the accuracy of essentially the most crucial binary job for the person: “Will I discover no less than one free port (Sure/No)?”
