# Decision-Focused Learning

In battery arbitrage, when a battery starts and ends the day at the same charge, a common shift in prices affects its schedule only through round-trip losses, so decision-focused losses identify the forecast level only weakly; clipping forecasts to a price cap can lower prediction error while removing profitable spreads. On 1,158 holdout days of the Indian day-ahead market, an equal-weight average of three statistical forecasts did better than the five decision-focused variants tested, and on 53 later heavily capped days no variant beat the regression or the average.

In battery arbitrage, when a battery starts and ends the day at the same charge, a common shift in prices affects its schedule only through round-trip losses, so decision-focused losses identify the forecast level only weakly; clipping forecasts to a price cap can lower prediction error while removing profitable spreads. On 1,158 holdout days of the Indian day-ahead market, an equal-weight average of three statistical forecasts did better than the five decision-focused variants tested, and on 53 later heavily capped days no variant beat the regression or the average.

Power swings at AI data centers make grid-import smoothing hard, and with limited battery power and energy, forecast errors of similar size can lead to very different scheduling costs. BS-DFL writes the battery scheduling objective into training: a constraint-aware neural optimization layer first learns schedules without optimal-solution labels, then is frozen so scheduling costs and ramping risk train the demand predictor, and deployment needs no optimization solver. Over a 10,000-minute simulation using loads derived from measured AI workloads, BS-DFL cut the mean absolute minute-to-minute grid-import change by 71.70% and total scheduling cost by 18.01% relative to an MSE-trained predictor with a Gurobi solver, at an average decision time of 18.77 ms.

Power swings at AI data centers make grid-import smoothing hard, and with limited battery power and energy, forecast errors of similar size can lead to very different scheduling costs. BS-DFL writes the battery scheduling objective into training: a constraint-aware neural optimization layer first learns schedules without optimal-solution labels, then is frozen so scheduling costs and ramping risk train the demand predictor, and deployment needs no optimization solver. Over a 10,000-minute simulation using loads derived from measured AI workloads, BS-DFL cut the mean absolute minute-to-minute grid-import change by 71.70% and total scheduling cost by 18.01% relative to an MSE-trained predictor with a Gurobi solver, at an average decision time of 18.77 ms.

Misclassifications from power-grid intrusion detection are converted into generator-level cyber-physical risk costs that enter dispatch optimization, instead of optimizing detection F1 alone. Previously detection and dispatch were optimized separately, so the operational cost of detection errors never entered the decision objective. On the ICS and Sherlock grid datasets, after a grid search over false-positive, false-negative and risk-scaling parameters, some configurations produced lower minimum dispatch costs than their baselines, with the size of the improvement varying by dataset and configuration.

Misclassifications from power-grid intrusion detection are converted into generator-level cyber-physical risk costs that enter dispatch optimization, instead of optimizing detection F1 alone. Previously detection and dispatch were optimized separately, so the operational cost of detection errors never entered the decision objective. On the ICS and Sherlock grid datasets, after a grid search over false-positive, false-negative and risk-scaling parameters, some configurations produced lower minimum dispatch costs than their baselines, with the size of the improvement varying by dataset and configuration.