Analysis of the Effectiveness of the Non-Vaccine Countermeasures Taken by the Indian Government against COVID-19 and Forecasting Using Machine Learning and Deep Learning

Shah, Akash; Shah, Romil; Gandhi, Manan; Panchani, Rashmil; Thakur, Govind; Srivastava, Kriti

doi:10.1201/9781003133681-22

Cited by 2 publications

(6 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We used default robot controllers to evaluate the learned policies. We used HARP (Shah and Srivastava 2022) with ϵ = 0 for computing zero-shot option guides.…”

Section: Resultsmentioning

confidence: 99%

“…Formally, a state abstraction from a concrete state space X to a finite, discrete state space S is a function that maps each x ∈ X to an element of S. State abstractions created by domain experts have been used extensively to speed up planning and learning. Recent work also presents approaches for learning state abstractions (e.g., (Konidaris, Kaelbling, and Lozano-Perez 2018;Shah and Srivastava 2022)). In this work we utilize the critical-region based state abstraction technique developed by Shah and Srivastava (2022) since it allows zeroshot state abstractions for new problems.…”

Section: State Abstractionsmentioning

confidence: 99%

“…Recent work also presents approaches for learning state abstractions (e.g., (Konidaris, Kaelbling, and Lozano-Perez 2018;Shah and Srivastava 2022)). In this work we utilize the critical-region based state abstraction technique developed by Shah and Srivastava (2022) since it allows zeroshot state abstractions for new problems. We present here a brief summary of this approach to highlight the properties and input requirements that are utilized in the current paper.…”

Section: State Abstractionsmentioning

confidence: 99%

“…Critical regions predictor We must first identify critical regions in an environment in order to synthesize state abstractions as discussed above. Shah and Srivastava (2022) provide an approach for learning generalizable critical region predictors in the form of a deep neural network. These predictors are environment independent and generalizable across robots to a large extent.…”

Section: State Abstractionsmentioning

confidence: 99%

“…E.g., the non-holonomic robots used to evaluate our approach (details in Sec. 5) are different from the robots used by Shah and Srivastava (2022), however, we used the critical regions predictor developed and made available by them for a rectangular holonomic robot without any additional training. We now briefly highlight the training process.…”

Section: State Abstractionsmentioning

confidence: 99%

See 4 more Smart Citations

Hierarchical Planning and Learning for Robots in Stochastic Settings Using Zero-Shot Option Invention

Shah,

Srivastava

2024

AAAI

View full text Add to dashboard Cite

This paper addresses the problem of inventing and using hierarchical representations for stochastic robot-planning problems. Rather than using hand-coded state or action representations as input, it presents new methods for learning how to create a high-level action representation for long-horizon, sparse reward robot planning problems in stochastic settings with unknown dynamics. After training, this system yields a robot-specific but environment independent planning system. Given new problem instances in unseen stochastic environments, it first creates zero-shot options (without any experience on the new environment) with dense pseudo-rewards and then uses them to solve the input problem in a hierarchical planning and refinement process. Theoretical results identify sufficient conditions for completeness of the presented approach. Extensive empirical analysis shows that even in settings that go beyond these sufficient conditions, this approach convincingly outperforms baselines by 2x in terms of solution time with orders of magnitude improvement in solution quality.

show abstract

“…We used default robot controllers to evaluate the learned policies. We used HARP (Shah and Srivastava 2022) with ϵ = 0 for computing zero-shot option guides.…”

Section: Resultsmentioning

confidence: 99%

Section: State Abstractionsmentioning

confidence: 99%

Section: State Abstractionsmentioning

confidence: 99%

Section: State Abstractionsmentioning

confidence: 99%

Section: State Abstractionsmentioning

confidence: 99%

See 3 more Smart Citations

Hierarchical Planning and Learning for Robots in Stochastic Settings Using Zero-Shot Option Invention

Shah,

Srivastava

2024

AAAI

View full text Add to dashboard Cite

show abstract

Learning Neuro-Symbolic Abstractions for Robot Planning and Learning

Shah

2024

AAAI

View full text Add to dashboard Cite

Although state-of-the-art hierarchical robot planning algorithms allow robots to efficiently compute long-horizon motion plans for achieving user desired tasks, these methods typically rely upon environment-dependent state and action abstractions that need to be hand-designed by experts. On the other hand, non-hierarchical robot planning approaches fail to compute solutions for complex tasks that require reasoning over a long horizon. My research addresses these problems by proposing an approach for learning abstractions and developing hierarchical planners that efficiently use learned abstractions to boost robot planning performance and provide strong guarantees of reliability.

show abstract

Analysis of the Effectiveness of the Non-Vaccine Countermeasures Taken by the Indian Government against COVID-19 and Forecasting Using Machine Learning and Deep Learning

Cited by 2 publications

References 0 publications

Hierarchical Planning and Learning for Robots in Stochastic Settings Using Zero-Shot Option Invention

Hierarchical Planning and Learning for Robots in Stochastic Settings Using Zero-Shot Option Invention

Learning Neuro-Symbolic Abstractions for Robot Planning and Learning

Contact Info

Product

Resources

About