Machine learning (ML) prediction APIs are increasingly widely used. An ML API can change over time due to model updates or retraining. This presents a key challenge in the usage of the API because it is often not clear to the user if and how the ML m
odel has changed. Model shifts can affect downstream application performance and also create oversight issues (e.g. if consistency is desired). In this paper, we initiate a systematic investigation of ML API shifts. We first quantify the performance shifts from 2020 to 2021 of popular ML APIs from Google, Microsoft, Amazon, and others on a variety of datasets. We identified significant model shifts in 12 out of 36 cases we investigated. Interestingly, we found several datasets where the APIs predictions became significantly worse over time. This motivated us to formulate the API shift assessment problem at a more fine-grained level as estimating how the API models confusion matrix changes over time when the data distribution is constant. Monitoring confusion matrix shifts using standard random sampling can require a large number of samples, which is expensive as each API call costs a fee. We propose a principled adaptive sampling algorithm, MASA, to efficiently estimate confusion matrix shifts. MASA can accurately estimate the confusion matrix shifts in commercial ML APIs using up to 90% fewer samples compared to random sampling. This work establishes ML API shifts as an important problem to study and provides a cost-effective approach to monitor such shifts.
Realization of strong optomechanical coupling in the single-photon level is crucial to study quantum nonlinear effects and manipulate macroscopic object. Here, we propose an alternative method to towards this goal in a hybrid ensemble-optomechanical
system. The sizable membrane-ensemble (ME) coupling mediated by the auxiliary mode of the cavity gives rise to polaritons with lower and higher frequencies. By tuning the ME coupling ($lambda_{rm en}$) approaching the critical coupling value ($lambda_c$), the eigen-energy of the low-frequency polariton gives rise to critical behavior (i.e., quantum phase transition) when the ensemble is within or beyond the low-excitation approximations. Using this critical behavior, the single-photon optomechanical coupling between the cavity and the low-frequency polariton can be greatly enhanced by almost three orders of magnitude with feasible parameters, while the coupling between the high-frequency polariton and the cavity is fully decouped. Our proposal provides a novel way to investigating Kerr effect and blockade in single-photon optomechanical systems.
Multi-label classification tasks such as OCR and multi-object recognition are a major focus of the growing machine learning as a service industry. While many multi-label prediction APIs are available, it is challenging for users to decide which API t
o use for their own data and budget, due to the heterogeneity in those APIs price and performance. Recent work shows how to select from single-label prediction APIs. However the computation complexity of the previous approach is exponential in the number of labels and hence is not suitable for settings like OCR. In this work, we propose FrugalMCT, a principled framework that adaptively selects the APIs to use for different data in an online fashion while respecting users budget. The API selection problem is cast as an integer linear program, which we show has a special structure that we leverage to develop an efficient online API selector with strong performance guarantees. We conduct systematic experiments using ML APIs from Google, Microsoft, Amazon, IBM, Tencent and other providers for tasks including multi-label image classification, scene text recognition and named entity recognition. Across diverse tasks, FrugalMCT can achieve over 90% cost reduction while matching the accuracy of the best single API, or up to 8% better accuracy while matching the best APIs cost.
As power systems are undergoing a significant transformation with more uncertainties, less inertia and closer to operation limits, there is increasing risk of large outages. Thus, there is an imperative need to enhance grid emergency control to maint
ain system reliability and security. Towards this end, great progress has been made in developing deep reinforcement learning (DRL) based grid control solutions in recent years. However, existing DRL-based solutions have two main limitations: 1) they cannot handle well with a wide range of grid operation conditions, system parameters, and contingencies; 2) they generally lack the ability to fast adapt to new grid operation conditions, system parameters, and contingencies, limiting their applicability for real-world applications. In this paper, we mitigate these limitations by developing a novel deep meta reinforcement learning (DMRL) algorithm. The DMRL combines the meta strategy optimization together with DRL, and trains policies modulated by a latent space that can quickly adapt to new scenarios. We test the developed DMRL algorithm on the IEEE 300-bus system. We demonstrate fast adaptation of the meta-trained DRL polices with latent variables to new operating conditions and scenarios using the proposed method and achieve superior performance compared to the state-of-the-art DRL and model predictive control (MPC) methods.
The multi-UAV network is promising to extend conventional networks by providing broader coverage and better reliability. Nevertheless, the broadcast nature of wireless signals and the broader coverage expose multi-UAV communications to the threats of
passive eavesdroppers. Recent studies mainly focus on securing a single legitimate link, or communications between a UAV and multiple ground users in one/two-UAV-aided networks, while the physical layer secrecy analysis for hierarchical multi-UAV networks is underexplored. In this paper, we investigate a general two-tier UAV network consisting of multiple UAV transmitters (UTs) and multiple UAV receivers (URs) in the presence of multiple UAV eavesdroppers (UEs). To protect all legitimate UT-UR links against UEs at the physical layer, we design a two-stage framework consisting of a UT-UR association stage and a cooperative transmission stage. Specifically, we formulate the secure transmission problem into a many-to-one matching game followed by an overlapping coalition formation (OCF) game, taking into account the limited capabilities and the throughput requirements of URs, as well as the transmission power constraints of UTs. A matching algorithm and an OCF algorithm are proposed to solve these two sequential games whose convergences and stabilities are guaranteed theoretically. Simulation results show the superiority of our algorithms and the effectiveness of our two-stage game framework in the terms of secrecy performance.
Load shedding has been one of the most widely used and effective emergency control approaches against voltage instability. With increased uncertainties and rapidly changing operational conditions in power systems, existing methods have outstanding is
sues in terms of either speed, adaptiveness, or scalability. Deep reinforcement learning (DRL) was regarded and adopted as a promising approach for fast and adaptive grid stability control in recent years. However, existing DRL algorithms show two outstanding issues when being applied to power system control problems: 1) computational inefficiency that requires extensive training and tuning time; and 2) poor scalability making it difficult to scale to high dimensional control problems. To overcome these issues, an accelerated DRL algorithm named PARS was developed and tailored for power system voltage stability control via load shedding. PARS features high scalability and is easy to tune with only five main hyperparameters. The method was tested on both the IEEE 39-bus and IEEE 300-bus systems, and the latter is by far the largest scale for such a study. Test results show that, compared to other methods including model-predictive control (MPC) and proximal policy optimization(PPO) methods, PARS shows better computational efficiency (faster convergence), more robustness in learning, excellent scalability and generalization capability.
Autonomous navigation is an essential capability of smart mobility for mobile robots. Traditional methods must have the environment map to plan a collision-free path in workspace. Deep reinforcement learning (DRL) is a promising technique to realize
the autonomous navigation task without a map, with which deep neural network can fit the mapping from observation to reasonable action through explorations. It should not only memorize the trained target, but more importantly, the planner can reason out the unseen goal. We proposed a new motion planner based on deep reinforcement learning that can arrive at new targets that have not been trained before in the indoor environment with RGB image and odometry only. The model has a structure of stacked Long Short-Term memory (LSTM). Finally, experiments were implemented in both simulated and real environments. The source code is available: https://github.com/marooncn/navbot.
