More Like this

1. Steady-State Convergence of Discrete Time State-Space Modeling with State-Dependent Switching

   https://doi.org/10.1109/COMPEL49091.2020.9265734  

   Baxter, Jared A. ; Costinett, Daniel J. ( November 2020 , 2020 IEEE 21st Workshop on Control and Modeling for Power Electronics (COMPEL))

   null (Ed.)

   Steady-state modeling plays an important role in the design of advanced power converters. Typically, steady-state modeling is completed by time-stepping simulators, which may be slow to converge to steady-state, or by dedicated analysis, which is time-consuming to develop across multiple topologies. Discrete time state-space modeling is a uniform approach to rapidly simulate arbitrary power converter designs. However, the approach requires modification to capture state-dependent switching, such as diode switching or current programmed modulation. This work provides a framework to identify and correct state-dependent switching within discrete time state-space modeling and shows the utility of the proposed method within the power converter design process. 

   more » « less
2. Broadening access to computing education state by state

   https://doi.org/10.1145/2856455  

   Adrion, Rick ; Fall, Renee ; Ericson, Barbara ; Guzdial, Mark ( January 2016 , Communications of the ACM)
3. The Illusion of State in State-Space Models

   Merrill, William ; Petty, Jackson ; Sabharwal, Ashish ( July 2024 , International Conference on Machine Learning (ICML) 2024)

   State-space models (SSMs) have emerged as a potential alternative architecture for building large language models (LLMs) compared to the previously ubiquitous transformer architecture. One theoretical weakness of transformers is that they cannot express certain kinds of sequential computation and state tracking (Merrill & Sabharwal, 2023), which SSMs are explicitly designed to address via their close architectural similarity to recurrent neural networks (RNNs). But do SSMs truly have an advantage (over transformers) in expressive power for state tracking? Surprisingly, the answer is no. Our analysis reveals that the expressive power of SSMs is limited very similarly to transformers: SSMs cannot express computation outside the complexity class 𝖳𝖢0. In particular, this means they cannot solve simple state-tracking problems like permutation composition. It follows that SSMs are provably unable to accurately track chess moves with certain notation, evaluate code, or track entities in a long narrative. To supplement our formal analysis, we report experiments showing that Mamba-style SSMs indeed struggle with state tracking. Thus, despite its recurrent formulation, the "state" in an SSM is an illusion: SSMs have similar expressiveness limitations to non-recurrent models like transformers, which may fundamentally limit their ability to solve real-world state-tracking problems. 

   more » « less
4. DriftSurf: Stable-State / Reactive-State Learning under Concept Drift

   Tahmasbi, Ashraf ; Jothimurugesan, Ellango ; Tirthapura, Srikanta ; Gibbons, Phillip ( July 2021 , Proceedings of the 38th International Conference on Machine Learning, ICML'21)
5. DriftSurf: Stable-State / Reactive-State Learning under Concept Drift

   Tahmasbi, Ashraf ; Jothimurugesan, Ellango ; Tirthapura, Srikanta ; Gibbons, Phillip B. ( July 2021 , Proceedings of the 38th International Conference on Machine Learning, {ICML} 2021, 18-24 July 2021, Virtual Even)

   When learning from streaming data, a change in the data distribution, also known as concept drift, can render a previously-learned model inaccurate and require training a new model. We present an adaptive learning algorithm that extends previous drift-detection-based methods by incorporating drift detection into a broader stable-state/reactive-state process. The advantage of our approach is that we can use aggressive drift detection in the stable state to achieve a high detection rate, but mitigate the false positive rate of standalone drift detection via a reactive state that reacts quickly to true drifts while eliminating most false positives. The algorithm is generic in its base learner and can be applied across a variety of supervised learning problems. Our theoretical analysis shows that the risk of the algorithm is (i) statistically better than standalone drift detection and (ii) competitive to an algorithm with oracle knowledge of when (abrupt) drifts occur. Experiments on synthetic and real datasets with concept drifts confirm our theoretical analysis. 

   more » « less